From jan at uran.informatik.uni-bonn.de Mon Feb 3 11:21:44 1997 From: jan at uran.informatik.uni-bonn.de (Jan Puzicha) Date: Mon, 3 Feb 1997 17:21:44 +0100 (MET) Subject: Preprints and Abstracts available online Message-ID: <199702031621.RAA02383@thalia.informatik.uni-bonn.de> This message has been posted to several lists. Sorry, if you receive multiple copies. The following six PREPRINTS are now available as abstracts and compressed postscript online via the WWW-Home-Page http://www-dbv.cs.uni-bonn.de/ of the |---------------------------------------------| |Computer Vision and Pattern Recognition Group| | of the University of Bonn, | | Prof. J. Buhmann, Germany. | |---------------------------------------------| 1.) Thomas Hofmann and Joachim Buhmann, Pairwise Data Clustering by Deterministic Annealing, IEEE Transactions on Pattern Analysis and Machine Intelligence PAMI 19(1), 1997. 2.) Joachim Buhmann and Thomas Hofmann, Robust Vector Quantization by Competitive Learning. In: Proceedings of the International Conference on Accoustics, Speech and Signal Processing (ICASSP'97), Munich, 1997. 3.) Hansjrg Klock and Joachim Buhmann, Multidimensional Scaling by Deterministic Annealing. In: Proceedings of the International Workshop on Energy Minimization Methods in Computer Vision and Pattern Recognition EMMCVPR'97, Venice. 4.) Thorsten Frhlinghaus and Joachim Buhmann. Real-Time Phase-Based Stereo for a Mobile Robot. in: Proceedings of the First Euromicro Workshop on Advanced Mobile robots. pp. 178-185, 1996. 5.) Thomas Hofmann, Jan Puzicha and Joachim M. Buhmann, Deterministic Annealing for Unsupervised Texture Segmentation. In: Proceedings of the International Workshop on Energy Minimization Methods in Computer Vision and Pattern Recognition (EMMCVPR'97), Venice. 6.) Thomas Hofmann, Jan Puzicha and Joachim M. Buhmann, A Deterministic Annealing Framework for Unsupervised Texture Segmentation. Technical Report IAI-TR-96-2, Institut fr Informatik III, University of Bonn. 1996. |--------------------------------------| | NEW: Test our algorithms on YOUR data| |--------------------------------------| If you want to test our unsupervised segmentation algorithms (see preprints 5. and 6. for more information) on your data, please download your data in any of the usual formats (pgm,gif,postscript,...) on our ftp-server uran.cs.uni-bonn.de in the directory /pub/dbv/incoming and write a short notice to jan at cs.uni-bonn.de. We will segment your data as soon as possible and send you the results. Feel free to have a look at the new online-presentation of texture segmentation results and the new links to related sites, conferences and jounals. If you have any questions or remarks, please let me know. Greetings Jan Puzicha -------------------------------------------------------------------- Jan Puzicha | email: jan at uran.cs.uni-bonn.de Institute f. Informatics III | jan at cs.uni-bonn.de University of Bonn | WWW : http://www.cs.uni-bonn.de/~jan | Roemerstrasse 164 | Tel. : +49 228 550-383 D-53117 Bonn | Fax : +49 228 550-382 -------------------------------------------------------------------- From rolf at rug104.cs.rug.nl Tue Feb 4 05:19:12 1997 From: rolf at rug104.cs.rug.nl (Rolf P. Wuertz) Date: Tue, 4 Feb 1997 11:19:12 +0100 (MET) Subject: New articles available from my website Message-ID: <199702041019.LAA02959@rug104.cs.rug.nl> The following articles are available online: --------------------------------------------------------------------- Rolf P. W"urtz. Neuronal theories and technical systems for face recognition. In Proceedings of the Fifth European Symposium On Artificial Neural Networks, Bruges (Belgium), 16-18 April 1997 ABSTRACT: I present various systems for the recognition of human faces. They consist of three steps: feature extraction, solving the correspondence problem, and the actual comparison with stored faces. Two of them are implemented in the Dynamic Link Architecture and are, therefore, close to biological hardware, the others are more technical in nature but also have some biological plausibility. At the end, I will briefly discuss the coherence with the results of psychophysical experiments on human face recognition. compressed PS source (350kB, 6 pages): http://www.cs.rug.nl/~rolf/esann97.ps.gz --------------------------------------------------------------------- Stephen McKenna, Sean Gong, Rolf P. W"urtz, Jonathan Tanner and Daniel Banin. Tracking facial feature points with Gabor wavelets and shape models. In Proceedings of the First International Conference on Audio- and Video-based Biometric Person Authentication Crans-Montana, Switzerland, 12-14 March 1997 ABSTRACT: A feature-based approach to tracking rigid and non-rigid facial motion is described. Feature points are characterised using Gabor Wavelets and can be individually tracked by phase-based displacement estimation. In order to achieve robust tracking a flexible shape model is used to impose global constraints upon the local feature points and to constrain the tracker. While there are many applications in facial analysis, the approach is quite generic and can be used for tracking other textured objects. compressed PS source (600 kB, 8 pages) http://www.cs.rug.nl/~rolf/gwt-pdm.ps.gz --------------------------------------------------------------------- Rolf P. W"urtz, Wolfgang Konen, and Kay-Ole Behrmann. How fast can neuronal algorithms match patterns? In Christoph von der Malsburg and Jan C. Vorbr"uggen and Werner von Seelen and Bernhard Sendhoff (eds.), Artificial Neural Networks - ICANN 96 , pages 145-150. Springer Verlag, 1996. ABSTRACT: We investigate the convergence speed of the Self Organizing Map (SOM) and Dynamic Link Matching (DLM) on a benchmark problem for the solution of which both algorithms are good candidates. We show that the SOM needs a large number of simple update steps and DLM a small number of complicated ones. A comparison of the actual number of floating point operations hints at an exponential vs. polynomial scaling behavior with increased pattern size. DLM turned out to be much less sensitive to parameter changes than the SOM. compressed PS source (60 kB, 6 pages) http://www.cs.rug.nl/~rolf/icann96.ps.gz +----------------------------------------------------------------------------+ | Rolf P. W"urtz | mailto: rolf at cs.rug.nl | URL: http://www.cs.rug.nl/~rolf/ | | Department of Computing Science, University of Groningen, The Netherlands | +----------------------------------------------------------------------------+ From wray at ptolemy-ethernet.arc.nasa.gov Wed Feb 5 01:12:06 1997 From: wray at ptolemy-ethernet.arc.nasa.gov (Wray Buntine) Date: Tue, 4 Feb 97 22:12:06 PST Subject: Research Scientist for autonomous data analysis Message-ID: <9702050612.AA01509@ptolemy.arc.nasa.gov> NASA's Center of Excellence in Information Technology at Ames Research Center invites candidates to apply for a position as Research Scientist in Information Technology: Position description: * We seek applicants to join a small team of space scientists and computer scientists in developing NASA's next generation smart spacecraft with on-board, autonomous data analysis systems. The group includes leading space scientists (Ted Roush, Virginia Gulick) and leading data analysts (Wray Buntine, Peter Cheeseman), and their counterparts at JPL. * The team is doing the research and development required for the task, and has a multi-year program with deliverables planned. This is not a pure research position, and requires dedication in seeing completion of the R&D milestones. * The applicant will be responsible for the information technology side of R&D, with guidance from senior space scientists on the project. * The research has strong links with on-going work at the Center of Excellence and is an integral part of NASA's long term goals. Candidate requirements: * Strong interest in demonstrating autonomous analysis systems to enhance science understanding in operational tests, with the ultimate goal of putting such systems in space. * Ph.D. degree in Computer Science, Electrical Engineering, or related field, and applied experience, possibly within the PhD. In exceptional cases, an M.S. degree with relevant work experience will suffice. * Knowledge of neural or probabilistic networks, machine learning, statistical pattern recognition, image processing, science data, processing, probabilistic algorithms, or related topics is essential. * Strong communication and organizational skills with the ability to lead a small team and interact with scientists. * Strong C programming and Unix skills (experimental, not necessarily production), with experience in programming mathematical algorithms: C++, Java, MatLab, IDL. Application deadline: * March 15th, 1997 (hardcopy required -- see below). Please send any questions by e-mail to the addresses below, and type "PI for Autonomous data analysis" as your header line. Dr. Ted Roush: roush at cloud9.arc.nasa.gov Dr. Wray Buntine: buntine at cloud9.arc.nasa.gov Full applications (which must include a resume and the names and addresses of at least two people familiar with your work) should be sent by surface mail (no e-mail, ftp or html applications will be accepted) to: Dr. Steve Lesh Attn: PI for Autonomous data analysis Mail Stop 269-1 NASA Ames Research Center Moffett Field, CA, 94025-1000 From hadley at cs.sfu.ca Wed Feb 5 18:14:16 1997 From: hadley at cs.sfu.ca (Bob Hadley) Date: Wed, 5 Feb 1997 15:14:16 -0800 (PST) Subject: an abstract to post Message-ID: <199702052314.PAA14631@melodious.cs.sfu.ca> Hi, I'd be grateful if someone could post this abstract on the Connectionist List. Thanks, -------------------------------------------------------------------- Robert F. Hadley (Bob) Phone: 604-291-4488 Associate Professor email: hadley at cs.sfu.ca School of Computing Science Simon Fraser University Burnaby, B.C. V5A 1S6 Canada ------------------------------------------------------------ FTP-host: ftp.fas.sfu.ca FTP-filename: /pub/cs/hadley/act.pas.ps The following paper is available by FTP and on the WWW. (instructions below) ------------------------------------- Acquisition of the Active-Passive Distinction from Sparse Input and No Error Feedback by Robert F. Hadley and Vlad C. Cardei School of Computing Science Simon Fraser University Burnaby, B.C. V5A 1S6 Canada hadley at cs.sfu.ca vcardei at cs.sfu.ca Technical Report -- CSS-IS TR 97-01 ABSTRACT A connectionist-inspired, parallel processing network is presented which learns, on the basis of (relevantly) sparse input, to assign meaning interpretations to *novel* test sentences in both active and passive voice. Training and test sentences are generated from a simple recursive grammar, but once trained, the network successfully processes thousands of sentences containing deeply embedded clauses. All training is unsupervised with regard to error feedback -- only Hebbian and Kohonen forms of training are employed. In addition, the active-passive distinction is acquired without any supervised provision of cues or flags (in the output layer) that indicate whether the input sentence is in active or passive sentence. In more detail: (1) The model learns on the basis of a corpus of about 1000 sentences while the set of potential test sentences contains over 100 million sentences. (2) The model generalizes its capacity to interpret active and passive sentences to substantially deeper levels of clausal embedding. (3) After training, the model satisfies criteria for strong syntactic and strong semantic *systematicity* that humans also satisfy. (4) Symbolic message passing occurs within the model's output layer. This symbolic aspect reflects certain prior *language acquistion* assumptions. (60 pages, 1.2 spacing) ----------------------------------------------------------- You can obtain the above paper via ftp by doing the following: ftp ftp.fas.sfu.ca When asked for your name, type the word: anonymous When asked for a password, use your e-mail address. Then, you should change directory as follows: cd pub/cs/hadley and then do a get, as in: get act.pas.ps To exit from ftp, type : quit .................. Also, available using web brousers at: http://fas.sfu.ca/cs/people/GradStudents/vcardei/personal/Projects/ From rybaki at eplrx7.es.dupont.com Wed Feb 5 15:20:07 1997 From: rybaki at eplrx7.es.dupont.com (Ilya Rybak) Date: Wed, 5 Feb 1997 15:20:07 -0500 Subject: information about WEB cites, please post Message-ID: <199702052020.PAA15815@pavlov> Dear colleagues, The WEB page of our models of respiratory pattern generation has been updated and now includes abstracts of recent publications. One of main objectives of this work was to consider integration of intrinsic cellular (e.g. kinetics of ionic channels , Ca dynamics) and network propitious , and the role of this integration in generation and control of the respiratory pattern. The address of the page is: http://www.voicenet.com/~rybak/resp.html The WEB page of BMV model (Behavioral Model of Visual perception and invariant recognition) has been also significantly updated. The model is developed on the basis of some principles of biological vision and is able to recognize complex images (scene objects, faces, etc.) invariantly in respect to shift, rotation and scale. A DEMO and recent paper may be downloaded from the page. The address is: http://www.voicenet.com/~rybak/vnc.html Interested people are welcome to visit both cites. (Please be patient, it takes some time to get the page first time). Best wishes, Ilya Rybak DuPont Central Research rybaki at eplrx7.es.dupont.com http://www.voicenet.com/~rybak/ From isbell at ai.mit.edu Thu Feb 6 10:21:17 1997 From: isbell at ai.mit.edu (Charles L Isbell) Date: Thu, 6 Feb 1997 10:21:17 -0500 (EST) Subject: MIMIC NIPS*96 paper available for ftp Message-ID: <199702061521.KAA25200@panther.ai.mit.edu> Recently, Shumeet Baluja announced on this list the availability of CMU-CS-97-107, "Using Optimal Dependency-Trees for Combinatorial Optimization: Learning the Structure of the Search Space" by Baluja and Davies. That paper discusses and extends some work presented in "MIMIC: Finding Optima by Estimating Probability Densities" by De Bonet, Isbell, and Viola (to appear in NIPS*96). It seems worthwhile to mention that this paper is also now available: ftp://ftp.ai.mit.edu/pub/isbell/mimic.ps.gz Abstract In many optimization problems, the structure of solutions reflects complex relationships between the different input parameters. For example, experience may tell us that certain parameters are closely related and should not be explored independently. Similarly, experience may establish that a subset of parameters must take on particular values. Any search of the cost landscape should take advantage of these relationships. We present MIMIC, a framework in which we analyze the global structure of the optimization landscape. A novel and efficient algorithm for the estimation of this structure is derived. We use knowledge of this structure to guide a randomized search through the solution space and, in turn, to refine our estimate of the structure. Our technique obtains significant speed gains over other randomized optimization procedures. Peace. From A.Sharkey at dcs.shef.ac.uk Thu Feb 6 05:49:56 1997 From: A.Sharkey at dcs.shef.ac.uk (Amanda Sharkey) Date: Thu, 6 Feb 97 10:49:56 GMT Subject: IEE Colloquium, Neural Networks for Industrial Applications Message-ID: <9702061049.AA25425@gw.dcs.shef.ac.uk> Announcing, IEE COLLOQUIUM: NEURAL NETWORKS FOR INDUSTRIAL APPLICATIONS. Wednesday February 12, 1997 at the Institution of Electrical Engineers, Savoy Place, London, WC2R OBL. Chairperson: Amanda Sharkey (University of Sheffield) Speakers: L. Tarassenko (University of Oxford, UK), 'Logicook and QUESTAR: two case studies in successful technology transfer' G. Hesketh (AEA Technology, plc, UK), 'COUNTERMATCH: a neural network approach to automatic signal verification' F. Fogelman Soulie (Sligos, France), 'Prediction in banking applications' H.H. Thodberg (Thodberg Scientific Computing, Denmark), 'The automation of intuitive visual expertise: a classification of beef carcasses using neural networks.' V. Tresp (Siemens AG, Germany), 'Neural Networks at Siemens: products and perspectives.' P. Cowley (Rolls-Royce plc, UK), 'Condition Monitoring high integrity systems.' A.J. Morris (University of Newcastle upon Tyne, UK) 'Building robust neural networks for industrial process.' Colloquium organised by Professional Group C4 (Artificial Intelligence) and co-sponsored by the British Computer Society Specialist Group on Expert Systems, the Society for the Study of Artificial Intelligence and Simulation of Human Behaviour (AISB), and the Neural Computing Applications Forum (NCAF). For information about registration, contact UK 0171 240 1871 Ext: 2205/6 From mkearns at research.att.com Thu Feb 6 15:25:58 1997 From: mkearns at research.att.com (Michael J. Kearns) Date: Thu, 6 Feb 1997 15:25:58 -0500 (EST) Subject: Paper on leave-one-out cross validation available Message-ID: <199702062025.PAA09030@radish.research.att.com> The following paper is now available in compressed postscript format at http://www.research.att.com/~mkearns under the "Model Selection/Complexity Regularization" section of my publications. Algorithmic Stability and Sanity-Check Bounds for Leave-One-Out Cross-Validation Michael Kearns (AT&T Labs) and Dana Ron (MIT) In this paper we prove "sanity-check" bounds for the error of the leave-one-out cross-validation estimate of the generalization error: that is, bounds showing that the worst-case error of this estimate is not much worse than that of the training error estimate. The name sanity-check refers to the fact that although we often expect the leave-one-out estimate to perform considerably better than the training error estimate, we are here only seeking assurance that its performance will not be considerably worse. Perhaps surprisingly, such assurance has been given only for rather limited cases in the prior literature on cross-validation. Any nontrivial bound on the error of leave-one-out must rely on some notion of algorithmic stability. Previous bounds relied on the rather strong notion of hypothesis stability, whose application was primarily limited to nearest-neighbor and other local algorithms. Here we introduce the new and weaker notion of error stability, and apply it to obtain sanity-check bounds for leave-one-out for other classes of learning algorithms, including training error minimization procedures and Bayesian algorithms. We also provide lower bounds demonstrating the necessity of error stability for proving bounds on the error of the leave-one-out estimate, and the fact that for training error minimization algorithms, in the worst case such bounds must still depend on the Vapnik-Chervonenkis dimension of the hypothesis class. From joachim at fit.qut.edu.au Fri Feb 7 00:48:56 1997 From: joachim at fit.qut.edu.au (Joachim Diederich) Date: Fri, 7 Feb 1997 15:48:56 +1000 (EST) Subject: PhD Scholarships Message-ID: <199702070548.PAA06756@aldebaran.fit.qut.edu.au> NEUROCOMPUTING RESEARCH CENTRE QUEENSLAND UNIVERSITY OF TECHNOLOGY (QUT) BRISBANE, AUSTRALIA FOUR PhD SCHOLARSHIPS QUT is offering PhD scholarships in the area of Neurocomputing and Artificial Intelligence. Research activities will include natural language acquisition, grammar induction and connectionist forms of analytical learning. The scholarships include an annual tax-free living allowance as well as the tuition fees for students from outside Australia/New Zealand. The successful applicants are expected to have a four-year degree in Computer Science/Cognitive Science (e.g. upper- levels Honours or Masters). Good knowledge of artificial neural networks and symbolic machine learning is essential; good programming skills are desirable. The scholars will work in the Neurocomputing Research Centre and will be part of a high-profile team comprising senior researchers and other PhD students. The centre offers state-of-the-art computing facilities including networked Unix workstations plus support for national and international conference travel. Potential applicants should contact Professor Joachim Diederich, Neurocomputing Research Centre, e-mail: joachim at fit.qut.edu.au. Enquiries should include a brief resume plus a reference to the scholarship of interest. These positions will be filled immediately. SCHOLARSHIP 1 NEUROCOMPUTING AND NATURAL LANGUAGE LEARNING QUT is offering a PhD scholarship in neurocomputing and natural language learning as part of a collaborative project with the University of California, Berkeley and the Australian National University. Research activities will include theoretical and experimental aspects of Connectionist Natural Language Processing, in particular, language learning. A strong background in computer science is required and qualifications in linguistics are an advantage. SCHOLARSHIP 2 COMPUTATIONAL INTELLIGENCE FOR THE PRIMARY INDUSTRIES This project investigates computerised techniques for the support of primary industies. In particular, statistical methods, neurocomputing and artificial intelligence techniques are evaluated. The comparison will be done by methods such as cross-validation and bootstrap, as well as by the use of "real world scenarios." The objective is an integrated information system which combines these techniques and aims at a performance which cannot be achieved by any of the above mentioned methods in isolation. SCHOLARSHIP 3 RECURRENT NEURAL NETWORKS AND CONTEXT-FREE LANGUAGES Various kinds of simple recurrent networks (SRNs) and their learning strategies will be evaluated and compared towards the implementation of a recogniser of context free languages modelled on a pushdown automaton. Starting with a Giles- style higher order network at the core, it is expected to devolve into a higher order recurrent network and a number of SRNs each with clearly delineated responsibilities. SCHOLARSHIP 4 CONNECTIONIST SYSTEMS AND ANALYTICAL LEARNING This project will focus on connectionists forms of analytical learning such as explanation-based generalisation. Connectionist systems for the representation of structured knowledge will be used for deduction and learning. Neurocomputing Research Centre Queensland University of Technology Box 2434, Brisbane Q 4001 AUSTRALIA Phone: +61 7 3864-2143 Fax: +61 7 3864-1801 http://www.fit.qut.edu.au/NRC/ From ucganlb at ucl.ac.uk Thu Feb 6 12:18:30 1997 From: ucganlb at ucl.ac.uk (Dr Neil Burgess - Anatomy UCL London) Date: Thu, 06 Feb 97 17:18:30 +0000 Subject: Discussion Meeting on Spatial Cognition, March 97 Message-ID: <33070.9702061718@link-1.ts.bcc.ac.uk> ``What are the parietal and hippocampal roles in spatial cognition?'' Date: 19th and 20th March 1997 Location: The Royal Society, 6 Carlton House Terrace, London SW1 5AG, U.K. Organisers: Neil Burgess & John O'Keefe Synopsis: The parietal cortex has long been identified as the neural substrate for spatial behaviour. More recently a spatial role for the hippocampus has been postulated, particularly in rats. Do these areas both perform the same computations, do they complement each other, are they both nodes in an overall network of spatial processing, or do they just have different functions in different species? These questions will be addressed with respect to neurophysiology, neuropsychology and computational modelling. Other Info: Royal Society discussion meetings are free to attend, but an attendance form should be returned to the Royal Society by the 10th March (lunch tickets cost 11.00 pounds sterling). Enquiries and requests for forms should be made to: Science Promotion Section, The Royal Society, 6 Carlton House Terrace, London SW1 5AG, U.K. Tel: +44 (0)171 839 5561 x2574, Fax: +44 (0)171 451 2693 Speakers: AMARAL, David (University of California) ANDERSEN, Richard (California Institute of Technology) ARBIB, Michael (University of Southern California) BERTHOZ, Alain (CNRS, Paris) BURGESS, Neil (University College London) GAFFAN, David (University of Oxford) KARNATH, Hans-Otto (University of Tubingen) MAGUIRE, Eleanor (Institute of Neurology, London) McNAUGHTON, Bruce (University of Arizona) MILNER, Brenda (Montreal Neurological Hospital) MISHKIN, Mortimer (NIMH, Bethesda) MORRIS, Richard (University of Edinburgh) MULLER, Robert (SUNY Health Sciences Center, Brooklyn) O'KEEFE, John (University College London) VALLAR, Guiseppe (University of Rome) Programme: Wednesday 19th March Session I. Chair: John O'Keefe (University College London) 9.45 O'KEEFE, John (University College London) Introduction 10.00 AMARAL, David (University of California) A neuroanatomical analysis of sensory inputs via temporal and parietal regions to the rat and monkey hippocampal formation 10.40 coffee 11.10 VALLAR, Guiseppe (University of Rome) Spatial frames of reference and somatosensory processing: a neuropsychological perspective 11.50 KARNATH, Hans-Otto (University of Tubingen) Spatial orientation and the representation of space with parietal lobe lesions 12.30 lunch Session II. Chair: Brenda Milner (Montreal Neurological Hospital) 2.00 ANDERSEN, Richard (California Institute of Technology) Multimodal integration for stable representations of space in the posterior parietal cortex 2.40 ARBIB, Michael (University of Southern California) Modeling the Specialization of Parietal Subregions in Providing Visual Affordances for Diverse Tasks 3.20 tea 3.50 BERTHOZ, Alain (CNRS, Paris) Neural basis of spatial memory during real and imagined locomotor trajectories 4.30 MISHKIN, Mortimer (NIMH, Bethesda) Cortical sensory processing streams and the hippocampus 5.10 general discussion 5.20 close Thursday 20th March Session III. Chair: Mortimer Mishkin (NIMH, Bethesda) 10.00 MILNER, Brenda (Montreal Neurological Hospital) Medial temporal lobe contributions to object location memory 10.40 coffee 11.10 MAGUIRE, Eleanor (Institute of Neurology, London) Hippocampal involvement in human topographical memory 11.50 GAFFAN, David (University of Oxford) Episodic memory, neglect and hemiamnesia 12.30 lunch Session IV. Chair: Lynn Nadel (University of Arizona) 2.00 MORRIS, Richard (University of Edinburgh) Hippocampal plasticity : synaptic changes underlying the automatic recording of attended experience 2.40 MULLER, Robert (SUNY Health Sciences Center, Brooklyn) A Topological approach to cognitive mapping 3.20 tea 3.50 McNAUGHTON, Bruce (University of Arizona) Memory reprocessing in cortico-cortico and hippocampo-cortical neuronal ensembles 4.30 BURGESS, Neil (University College London) Robotic and neuronal simulation of the hippocampus and rat navigation 5.10 general discussion 5.20 round-up 5.30 Close From wyler at iam.unibe.ch Fri Feb 7 11:01:39 1997 From: wyler at iam.unibe.ch (Kuno Wyler) Date: Fri, 7 Feb 1997 17:01:39 +0100 Subject: POSTDOCTORAL RESEARCH FELLOWSHIP available Message-ID: <9702071601.AA06468@garfield.unibe.ch> POSTDOCTORAL RESEARCH FELLOWSHIP -------------------------------- Neural Computing Research Group Institute of Informatics and Applied Mathematics University of Bern, Switzerland The Neural Computing Research Group at the University of Bern is looking for a highly motivated individual for a two year postdoctoral research position as project manager in the area of development of a neuromorphic perception system based on multi sensor fusion. The aim of the project is to develop a neurobiologically plausible perception system for novelty detection in a real world environment (e.g. quality control in industrial production lines or supervision of security zones) using the fusion of audio and video information and fast learning algorithms for the dynamic binding of signal features. Potential candidates should have good skills in applied optical sciences with a sound background in mathematics and signal processing, and working knowledge in neurobiology and neural networks. Further knowledge of programming (Matlab, LabView or C/C++) or VLSI technology is highly desirable but not required. Project management includes the supervision and the scientific leading of the collaborators (1 Postdoc/ 2 PhD candidates). The position will begin May 1, 1997, with possible renewal for an additional two years. The initial salary is SFr. 70'000/year (approx. $50'000). To apply for this position, send your curriculum vitae, publication list with one or two sample publications and two letters of reference before March 15, 1997, by surface mail to (applications by e-mail will not be accepted) Neural Computing Research Group Postdoctoral Research Fellowship attn. Mrs. M. Aberegg Institute of Informatics and Applied Mathematics University of Bern Neubrueckstrasse 10 CH-3012 Bern Switzerland For further information you can send e-mail to wyler at iam.unibe.ch or mueller at iam.unibe.ch. From reza at bme.jhu.edu Fri Feb 7 12:40:39 1997 From: reza at bme.jhu.edu (Reza Shadmehr) Date: Fri, 7 Feb 1997 12:40:39 -0500 (EST) Subject: faculty position in computational neuroscience Message-ID: <199702071740.MAA17245@reflex.bme.jhu.edu> JOHNS HOPKINS UNIVERSITY BIOMEDICAL ENGINEERING The Department of Biomedical Engineering at the Johns Hopkins University invites applications for a tenured or tenure-track faculty position in computational neuroscience. The successful candidate must be able to establish an independently funded research program that builds on or complements existing research programs. The candidate will also participate in the teaching of undergraduate and graduate students. For more information on our department see: http://www.bme.jhu.edu. Interested applicants should send their CV and names of three referees to: Dr. Murray B. Sachs Faculty Search Department of Biomedical Engineering 710 Traylor Building 720 Rutland Avenue Baltimore, MD 21205 The Johns Hopkins University is an Equal Opportunity/Affirmative Action Institution. From Dave_Touretzky at DST.BOLTZ.CS.CMU.EDU Sun Feb 9 21:44:50 1997 From: Dave_Touretzky at DST.BOLTZ.CS.CMU.EDU (Dave_Touretzky@DST.BOLTZ.CS.CMU.EDU) Date: Sun, 09 Feb 97 21:44:50 EST Subject: CNBC summer undergraduate research program Message-ID: <5114.855542690@DST.BOLTZ.CS.CMU.EDU> The Center for the Neural Basis of Cognition, a joint project of Carnegie Mellon University and the University of Pittsburgh, is seeking applications from top-quality undergraduates interested in pursuing summer research in cognitive or computational neuroscience. The CNBC summer training program is a ten week intensive program of lectures, laboratory tours, and guided research. State of the art facilities include computerized microscopy; laboratories for human and animal electrophysiological recording; behavioral assessment laboratories for rat, primate, and human experimentation; MRI and PET scanners for brain imaging; the Pittsburgh Supercomputing Center; and a regional medical center providing access to human clinical populations. The Summer Training Program is a National Science Foundation sponsored program; we expect to support ten students in each of the next five years. Applications are encouraged from students with interests in biology, neuroscience psychology, engineering, physics, mathematics, computer science, or robotics. To be eligible, students must not yet have completed their bachelor's degree at the time they participate. The application deadline is March 15, 1997. For more information about the program, and detailed application instructions, see our web site at: http://www.cnbc.cmu.edu/Training/summer/index.html In addition to its summer program, the Center for the Neural Basis of Cognition (CNBC) offers an interdisciplinary training program for Ph.D. and postdoctoral students in collaboration with various affiliated departments at Carnegie Mellon University and the University of Pittsburgh. This training program is the descendant of the Neural Processes in Cognition program started in 1990 under the National Science Foundation. We now have thirty six graduate students and thirty five faculty affiliated with the CNBC. The program focuses on understanding higher level brain function in terms of neurophysiological, cognitive, and brain imaging data complemented with computational modeling. Individually designed programs of study encompass cellular and systems neuroscience, computational neuroscience, cognitive modeling, and brain imaging. For a brochure describing the graduate training program and application materials, contact us at the following address: Center for the Neural Basis of Cognition 115 Mellon Institute 4400 Fifth Avenue Pittsburgh, PA 15213 Telephone: (412) 268-4000 Fax: (412) 268-5060 Email: cnbc-admissions at cnbc.cmu.edu This material is also available on our web site at http://www.cnbc.cmu.edu Faculty: The CNBC training faculty includes: German Barrionuevo (Pitt Neuroscience): LTP in hippocampal slice; Marlene Behrmann (CMU Psychology): spatial representations in parietal cortex; Pat Carpenter (CMU Psychology): mental imagery, language, and problem solving; Jonathan Cohen (CMU Psychology): schizophrenia; dopamine and attention; Carol Colby (Pitt Neuroscience): spatial reps. in primate parietal cortex; Bard Ermentrout (Pitt Mathematics): oscillations in neural systems; Julie Fiez (Pitt Psychology): fMRI studies of language; John Horn (Pitt Neurobiology): synaptic learning in autonomic ganglia; Allen Humphrey (Pitt Neurobiology): motion processing in primary visual cortex; Marcel Just (CMU Psychology): visual thinking, language comprehension; Eric Klann (Pitt Neuroscience): hippocampal LTP and LTD; Alan Koretsky (CMU Biological Sciences): new fMRI techniques for brain imaging; Tai Sing Lee (CMU Comp. Sci.): primate visual cortex; computer vision; David Lewis (Pitt Neuroscience): anatomy of frontal cortex; James McClelland (CMU Psychology): connectionist models of cognition; Carl Olson (CNBC): spatial representations in primate frontal cortex; David Plaut (CMU Psychology): connectionist models of reading; Michael Pogue-Geile (Pitt Psychology): development of schizophrenia; John Pollock (CMU Biological Sci.): neurodevelopment of the fly visual system; Walter Schneider (Pitt Psychology): fMRI studies of attention and skill acquisition; Charles Scudder (Pitt Neurobiology): motor learning in cerebellum; Susan Sesack (Pitt Neuroscience): anatomy of the dopaminergic system; Dan Simons (Pitt Neurobiology): sensory physiology of the cerebral cortex; William Skaggs (Pitt Neuroscience): representations in rodent hippocampus; and David Touretzky (CMU Comp. Sci.): hippocampus, rat navigation, animal learning. Walter Schneider David Touretzky Professor of Psychology Computer Science Department University of Pittsburgh Carnegie Mellon University From meunier at orphee.polytechnique.fr Mon Feb 10 08:02:44 1997 From: meunier at orphee.polytechnique.fr (Claude Meunier) Date: Mon, 10 Feb 1997 14:02:44 +0100 Subject: No subject Message-ID: Research associate Center for Theoretical Physics Ecole Polytechnique, France The Center for Theoretical Physics invites applications for a possible one year position of research associate (no teaching duties), starting Fall 1997. Six month stays may also be considered. Candidates are expected to carry out their research in the framework of our existing research programs in: - computational neurosciences - condensed matter - field theory and particle physics - plasma physics Applicants should submit a resume with publications list, statement of research interests, and two letters of references to: Dr. M.-N. Bussac Centre de Physique Th?orique Ecole Polytechnique 91128 Palaiseau cedex France Fax: 00 33 01 69 33 30 08 E-mail: bussac at cpth.polytechnique.fr Send requests of information on this temporary position and post doctoral positions in theoretical physics at Ecole Polytechnique to the same address. The closing date for the receipt of applications is March 15, 1997. From klaus at prosun.first.gmd.de Mon Feb 10 13:22:12 1997 From: klaus at prosun.first.gmd.de (Klaus-R. Mueller) Date: Mon, 10 Feb 1997 19:22:12 +0100 Subject: Call for Tricks ;-) Message-ID: <9702101822.AA06771@prosun.first.gmd.de> ********************************************* * Call for Papers * * Tricks of the Trade * ********************************************* Dear Colleagues, At the Nips*96 workshops we had a workshop called "Tricks of the Trade: How to Make Algorithms Really Work". As a follow-up to this workshop, we are collecting papers for a "Book of Tricks" (this is our working title), which will be tentatively published in the LNCS State-of-the-Art Surveys series. We would like to invite you to contribute. What is a Trick?: A technique, rule-of-thumb, or heuristic that * is easy to describe and understand * can make a real difference in practice * is not (yet) part of well documented technique * has broad application and may or may not (yet) have a theoretical explanation. Content and Format: In order to keep everything focussed, we suggest the following main topics of the book: 1. architectural tricks 2. sampling and data preprocessing 3. speeding learning procedures 4. improving generalization and optimization To give consistency across papers, we would like there to be a structural similarity in the contents which is based on the outline we proposed for the talks: - intro/motivation - trick - where has it been tried and how well does it work? - why does it work? - possible theoretical explanation (if any), - heuristic explanation (if any) & discussion For More Details: Please look at the call for papers on our web site: In the US: http://www.willamette.edu/~gorr/tricks/guidelines.html In Europe: http://www.first.gmd.de/persons/Mueller.Klaus-Robert/CALL.html If you have questions, please email to: Jenny Orr (gorr at willamette.edu) Klaus-Robert M"uller (klaus at first.gmd.de) Rich Caruana (caruana at cs.cmu.edu) From meunier at orphee.polytechnique.fr Tue Feb 11 04:02:01 1997 From: meunier at orphee.polytechnique.fr (Claude Meunier) Date: Tue, 11 Feb 1997 10:02:01 +0100 Subject: research associate position Message-ID: Research associate Center for Theoretical Physics Ecole Polytechnique, France The Center for Theoretical Physics invites applications for a possible one year position of research associate (no teaching duties), starting Fall 1997. Six month stays may also be considered. Candidates are expected to carry out their research in the framework of our existing research programs in: - computational neurosciences - condensed matter - field theory and particle physics - plasma physics Applicants should submit a resume with publications list, statement of research interests, and two letters of references to: Dr. M.-N. Bussac Centre de Physique Th=E9orique Ecole Polytechnique 91128 Palaiseau cedex France fax: 00 33 01 69 33 30 08 E-mail: bussac at cpth.polytechnique.fr Send requests of information on this temporary position and post doctoral positions in theoretical physics at Ecole Polytechnique to the same address. The closing date for the receipt of applications is March 15, 1997. From karaali at ukraine.corp.mot.com Tue Feb 11 12:34:16 1997 From: karaali at ukraine.corp.mot.com (Orhan Karaali) Date: Tue, 11 Feb 1997 11:34:16 -0600 Subject: Motorola NN Speech Synthesizer Article Message-ID: <199702111734.LAA03748@fiji.mot.com> FTP-host: archive.cis.ohio-state.edu FTP-filename: /pub/neuroprose/karaali.synthesis_wcnn96.ps.Z Motorola Neural Network Speech Synthesizer Article A new neural network based speech synthesizer has been developed here at Motorola Chicago Corporate Research Laboratories by the Speech Synthesis and Machine Learning Group. We believe that the quality of the synthesized speech it produces surpasses the current state of the art, particularly in naturalness. An invited paper describing this neural network speech synthesizer was presented in the Speech Session of the World Congress on Neural Networks 96 in San Diego. This paper is now available in the NEUROPROSE archive as karaali.synthesis_wcnn96.ps.Z. If you have a problem getting the paper from NEUROPROSE, I can email it to you. Orhan Karaali email: karaali at mot.com --------------------------------------------------------------------- Speech Synthesis with Neural Networks Orhan Karaali, Gerald Corrigan, and Ira Gerson Motorola, Inc., 1301 E. Algonquin Road, Schaumburg, IL 60196 karaali at mot.com, corrigan at mot.com, gerson at mot.com ABSTRACT Text-to-speech conversion has traditionally been performed either by concatenating short samples of speech or by using rule-based systems to convert a phonetic representation of speech into an acoustic representation, which is then converted into speech. This paper describes a system that uses a time-delay neural network (TDNN) to perform this phonetic-to-acoustic mapping, with another neural network to control the timing of the generated speech. The neural network system requires less memory than a concatenation system, and performed well in tests comparing it to commercial systems using other technologies. ----- End Included Message ----- From lawrence at research.nj.nec.com Wed Feb 12 13:05:43 1997 From: lawrence at research.nj.nec.com (Steve Lawrence) Date: Wed, 12 Feb 1997 13:05:43 -0500 Subject: Student Position: Learning for Agents and Multi-Agent Systems Message-ID: <19970212130543.19266@purgatory.nj.nec.com> The NEC Research Institute in Princeton, NJ has an immediate opening for a student research position in the area of learning for agents and multi-agent systems. Candidates must have experience in research and be able to effectively communicate research results. Ideal candidates will have knowledge of one or more machine learning techniques (e.g. neural networks, decision trees, rule based systems, and nearest neighbor techniques), and be proficient in the software implementation of algorithms. NEC Research provides an outstanding research environment with many recognized experts and excellent resources including several multiprocessor machines. Interested applicants should apply by email, mail or fax including their resumes and any specific interests in learning for agents and multi-agent systems to: Dr. C. Lee Giles NEC Research Institute 4 Independence Way Princeton NJ 08540 Phone: (609) 951 2642 Fax: (609) 951 2482 Email: giles at research.nj.nec.com -- Steve Lawrence <*> http://www.neci.nj.nec.com/homepages/lawrence From wermter at nats5.informatik.uni-hamburg.de Wed Feb 12 07:45:25 1997 From: wermter at nats5.informatik.uni-hamburg.de (Stefan Wermter) Date: Wed, 12 Feb 1997 13:45:25 +0100 Subject: JAIR article - connectionist natural language learning Message-ID: <199702121245.NAA12800@nats6> A non-text attachment was scrubbed... Name: not available Type: x-sun-attachment Size: 4021 bytes Desc: not available Url : https://mailman.srv.cs.cmu.edu/mailman/private/connectionists/attachments/00000000/c4d77933/attachment.ksh From beiu at lanl.gov Tue Feb 11 22:50:18 1997 From: beiu at lanl.gov (Valeriu Beiu) Date: Tue, 11 Feb 1997 20:50:18 -0700 Subject: CFP: IV Brazilian Symposium on Neural Networks SBRN'97 Message-ID: <2.2.32.19970212035018.003026f0@nis-pop.lanl.gov> /\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\ I V t h B R A Z I L I A N S Y M P O S I U M O N N E U R A L N E T W O R K S SBRN'97 /\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\ Goiania, GO, Brazil, December 03 - 05, 1997 Sponsored by: Brazilian Computer Society (SBC) International Neural Networks Society (INNS) Organized by the School of Electrical Engineering (EEE) Federal University of Goias (UFG) Conference Chairmen: Weber Martins (EEE-UFG) Dibio Leandro Borges (EEE-UFG) ======================================================================= Latest information will be placed at the conference WWW-page http://www.eee.ufg.br/sbrn97/ ======================================================================= P u r p o s e ----------------------------------------------------------------------- The fourth Brazilian Symposium on Neural Networks will be held at the School of Electrical Engineering, Federal University of Goias (UFG), Goiania, GO (Brazil), on 3rd, 4th, and 5th December, 1997. The Brazilian Symposia on Neural Networks are organized by the interest group in Neural Networks of the Brazilian Computer Society. In order to promote research in Neural Networks and Evolutionary Computation, and scientific interchange among Brazilian AI researchers and practitioners, and their counterparts worldwide, papers from the international community are most welcome. The papers will be reviewed by an international program committee. T o p i c s o f I n t e r e s t ----------------------------------------------------------------------- Submissions are invited on substantial, original, and previously unpublished research in all aspects of Neural Networks, including, but not limited to: * Biological Perspectives * Cognitive Modeling * Dynamic Systems * Evolutionary Computation * Fuzzy Logic * Hardware Implementation * Hybrid Systems * Learning Models * Neural Network Algorithms and Architectures * Neural Network Applications * Optimization * Pattern Recognition and Image Processing * Robotics and Control * Signal Processing * Theoretical Models P r o g r a m C o m m i t t e e ----------------------------------------------------------------------- Nigel M. Allinson (Univ. Manchester Institute of Technology - UK) Jose Nelson Amaral (Pontificia Catholic University, RS - BRAZIL) William Armstrong (University of Alberta - CANADA) Valeriu Beiu (Los Alamos National Laboratory - USA) Dibio Leandro Borges * (Federal University of Goias - BRAZIL) Antonio de Padua Braga (Federal University of Minas Gerais - BRAZIL) Andre L.P. de Carvalho (University of Sao Paulo, S. Carlos - BRAZIL) Weber Martins * (Federal University of Goias - BRAZIL) Edilberto P. Teixeira (Federal University of Uberlandia - BRAZIL) Harold Szu (University of Southwestern Louisiana - USA) Germano C. Vasconcelos (Federal University of Pernambuco - BRAZIL) * Program Chairmen T i m e t a b l e ----------------------------------------------------------------------- Submissions due May 2nd, 1997 Notification of acceptance July 5th, 1997 Final manuscript due August 5th, 1997 Conference date December 3rd-5th, 1997 P a p e r S u b m i s s i o n s ----------------------------------------------------------------------- We invite submissions of scientific papers to any topic related to Neural Networks and Evolutionary Computation. Authors should submit four (4) copies of their papers in hard copy form. Neither computer files nor fax submission are acceptable. Submissions must be printed on 8 1/2 x 11 inch (21.59 x 27.94 cm) or A4 paper using 12 point type, and they must be a maximum of twenty (20) pages long, double space, including all figures and references. The reviewing process will be blind to the identities of the authors, please note that this requires that authors exercise some care not to identify themselves in their papers. Title page and paper body ------------------------- Each copy of the paper must include a title page, separate from the body of the paper, containing the title of the paper, the names and addresses of all authors (please, complete addresses including also e-mail and FAX), a short abstract of less than 200 words, and a list of keywords giving the area/subarea of the paper. The second page, on which the paper body begins, should include the same title, abstract, keywords, but not the names and affiliations of the authors. All paper submissions should be to the following address: --------------------------------------------------------- IVth Brazilian Symposium on Neural Networks (SBRN'97) (PAPER SUBMISSION) School of Electrical Engineering (EEE) Federal University of Goias (UFG) Pca. Universitaria s/n Setor Universitario 74605-220 Goiania, GO (BRAZIL) FAX: +55 62 202 - 0325 E-mail: sbrn97 at eee.ufg.br Review criteria --------------- Papers will be subject to peer review by an international program committee. Selection criteria include accuracy and originality of ideas, clarity and significance of results and the quality of the presentation. P u b l i c a t i o n ----------------------------------------------------------------------- We are planning to publish the international proceedings with a major publishing house, and details will be available soon. Papers submitted in Portuguese or Spanish will be published in a separate volume. Both proceedings will be available at the conference. Please, note that at least one author should attend the conference for presentation of the accepted paper. I n q u i r i e s ----------------------------------------------------------------------- Inquiries regarding any aspect of the conference may be sent to the internet address: sbrn97 at eee.ufg.br or to the Conference Chairmen: Weber Martins and Dibio Leandro Borges both at: School of Electrical Engineering Federal University of Goias Pca. Universitaria s/n Setor Universitario 74605-220 Goiania, GO (BRAZIL) \/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\ From jbednar at cs.utexas.edu Fri Feb 14 17:11:31 1997 From: jbednar at cs.utexas.edu (jbednar@cs.utexas.edu) Date: Fri, 14 Feb 1997 16:11:31 -0600 Subject: Overview paper and theses on the RF-LISSOM project Message-ID: <199702142211.QAA35148@duckula.cs.utexas.edu> The following paper (to appear in Psychology of Learning and Motivation) gives an overview of the RF-LISSOM project on modeling the primary visual cortex, ongoing at the UTCS Neural Networks Research Group. For those seeking details, please refer to the dissertation and the theses also announced below. All these publications (and others) are available from our web page at http://www.cs.utexas.edu/users/nn (under publications, self-organization; the direct ftp addresses are included below as well). Public domain LISSOM code, for self-organization in laterally connected maps, will be announced in the near future. -- The Authors ------------------------------------------------------------------------- SELF-ORGANIZATION, PLASTICITY, AND LOW-LEVEL VISUAL PHENOMENA IN A LATERALLY CONNECTED MAP MODEL OF THE PRIMARY VISUAL CORTEX Risto Miikkulainen, James A. Bednar, Yoonsuck Choe, and Joseph Sirosh Department of Computer Sciences, The University of Texas at Austin. In R. L. Goldstone, P. G. Schyns, and D. L. Medin (eds.), Psychology of Learning and Motivation, vol. 36, 1997 in press (36 pages). ftp://ftp.cs.utexas.edu/pub/neural-nets/papers/miikkulainen.visual-cortex.ps.Z Based on a Hebbian adaptation process, the afferent and lateral connections in the RF-LISSOM model organize simultaneously and cooperatively, and form structures such as those observed in the primary visual cortex. The neurons in the model develop local receptive fields that are organized into orientation, ocular dominance, and size selectivity columns. At the same time, patterned lateral connections form between neurons that follow the receptive field organization. This structure is in a continuously-adapting dynamic equilibrium with the external and intrinsic input, and can account for reorganization of the adult cortex following retinal and cortical lesions. The same learning processes may be responsible for a number of low-level functional phenomena such as tilt aftereffects, and combined with the leaky integrator model of the spiking neuron, for segmentation and binding. The model can also be used to verify quantitatively the hypothesis that the visual cortex forms a sparse, redundancy-reduced encoding of the input, which allows it to process massive amounts of visual information efficiently. ------------------------------------------------------------------------- A SELF-ORGANIZING NEURAL NETWORK MODEL OF THE PRIMARY VISUAL CORTEX Joseph Sirosh Department of Computer Sciences, The University of Texas at Austin. PhD Dissertation and Technical Report AI95-237, August 1995 (137 pages). ftp://ftp.cs.utexas.edu/pub/neural-nets/papers/sirosh.diss.tar This work is aimed at modeling and analyzing the computational processes by which sensory information is learned and represented in the brain. First, a general self-organizing neural network architecture that forms efficient representations of visual inputs is presented. Two kinds of visual knowledge are stored in the cortical network: information about the principal feature dimensions of the visual world (such as line orientation and ocularity) is stored in the afferent connections, and correlations between these features in the lateral connections. During visual processing, the cortical network filters out these correlations, generating a redundancy-reduced sparse coding of the visual input. Through massively parallel computational simulations, this architecture is shown to give rise to structures similar to those in the primary visual cortex, such as (1) receptive fields, (2) topographic maps, (3) ocular dominance, orientation and size preference columns, and (4) patterned lateral connections between neurons. The same computational process is shown to account for many of the dynamic processes in the visual cortex, such as reorganization following retinal and cortical lesions, and perceptual shifts following dynamic receptive field changes. These results suggest that a single self-organizing process underlies development, plasticity and visual functions in the primary visual cortex. ------------------------------------------------------------------------- TILT AFTEREFFECTS IN A SELF-ORGANIZING MODEL OF THE PRIMARY VISUAL CORTEX James A. Bednar Department of Computer Sciences, The University of Texas at Austin. Masters Thesis and Technical Report AI97-259, January 1997 (104 pages). ftp://ftp.cs.utexas.edu/pub/neural-nets/papers/bednar.thesis.tar The psychological phenomenon known as the tilt aftereffect was used to demonstrate the functional properties of RF-LISSOM, a self-organizing model of laterally connected orientation maps in the primary visual cortex. The same self-organizing processes that are responsible for the development of the map and its lateral connections are shown to result in tilt aftereffects as well. The model allows analysis of data that are difficult to measure in humans, thus providing a view of the cortex that is otherwise not available. The results give computational support for the idea that tilt aftereffects arise from lateral interactions between adapting feature detectors, as has long been surmised. They also suggest that indirect tilt aftereffects could result from the conservation of synaptic resources. The model thus provides a unified computational explanation of self-organization and both direct and indirect tilt aftereffects in the primary visual cortex. ------------------------------------------------------------------------- LATERALLY INTERCONNECTED SELF-ORGANIZING FEATURE MAP IN HANDWRITTEN DIGIT RECOGNITION Yoonsuck Choe Department of Computer Sciences, The University of Texas at Austin. Masters Thesis and Technical Report AI95-236, August 1995 (65 pages). ftp://ftp.cs.utexas.edu/pub/neural-nets/papers/choe.thesis.tar.Z An application of biologically motivated laterally interconnected synergetically self-organizing maps (LISSOM) to off-line recognition of handwritten digit is presented. The lateral connections of the LISSOM map learn the correlations between units through Hebbian learning. As a result, the excitatory connections focus the activity in local patches and lateral connections decorrelate redundant activity on the map. This process forms internal representations for the input that are easier to recognize than the input bitmaps themselves or the activation patterns on a standard Self-Organizing Map (SOM). The recognition rate on a publically available subset of NIST special database 3 with LISSOM is 4.0% higher than that based on SOM, and 15.8% higher than that based on raw input bitmaps. These results form a promising starting point for building pattern recognition systems with a LISSOM map as a front end. From harnad at cogsci.soton.ac.uk Sun Feb 16 09:44:27 1997 From: harnad at cogsci.soton.ac.uk (Stevan Harnad) Date: Sun, 16 Feb 97 14:44:27 GMT Subject: Feature Creation: BBS Call for Commentators Message-ID: <20512.9702161444@cogsci.ecs.soton.ac.uk> Below is the abstract of a forthcoming BBS target article on: THE DEVELOPMENT OF FEATURES IN OBJECT CONCEPTS by Philippe G. Schyns, Robert L. Goldstone & Jean-Pierre Thibaut This article has been accepted for publication in Behavioral and Brain Sciences (BBS), an international, interdisciplinary journal providing Open Peer Commentary on important and controversial current research in the biobehavioral and cognitive sciences. Commentators must be BBS Associates or nominated by a BBS Associate. To be considered as a commentator for this article, to suggest other appropriate commentators, or for information about how to become a BBS Associate, please send EMAIL to: bbs at cogsci.soton.ac.uk or write to: Behavioral and Brain Sciences Department of Psychology University of Southampton Highfield, Southampton SO17 1BJ UNITED KINGDOM http://www.princeton.edu/~harnad/bbs/ http://www.cogsci.soton.ac.uk/bbs/ ftp://ftp.princeton.edu/pub/harnad/BBS/ ftp://ftp.cogsci.soton.ac.uk/pub/bbs/ gopher://gopher.princeton.edu:70/11/.libraries/.pujournals If you are not a BBS Associate, please send your CV and the name of a BBS Associate (there are currently over 10,000 worldwide) who is familiar with your work. All past BBS authors, referees and commentators are eligible to become BBS Associates. To help us put together a balanced list of commentators, please give some indication of the aspects of the topic on which you would bring your areas of expertise to bear if you were selected as a commentator. An electronic draft of the full text is available for inspection with a WWW browser, anonymous ftp or gopher according to the instructions that follow after the abstract. ____________________________________________________________________ THE DEVELOPMENT OF FEATURES IN OBJECT CONCEPTS Philippe G. Schyns Robert L. Goldstone Jean-Pierre Thibaut Psychology Dept. Psychology Dept. Psychology Dept. Glasgow University Indiana University Universite de Liege Glasgow G12 8QB UK Bloomington IN 47405 4000 Liege BELGIUM philippe at psy.gla.ac.uk rgoldsto at ucs.indiana.edu jthibaut at vm1.ulg.ac.be KEYWORDS: Concept learning, conceptual development, perceptual learning, features, stimulus encoding ABSTRACT: According to an influential approach to cognition, our perceptual systems provide us with a repertoire of fixed features as input to higher-level cognitive processes. We present a theory of category learning and representation in which features, instead of being components of a fixed repertoire, are created under the influence of higher-level cognitive processes. When new categories need to be learned, fixed features face one of two problems: (1) High-level features that are directly useful for categorization may not be flexible enough to represent all relevant objects. (2) Low-level features consisting of unstructured fragments (such as pixels) may not capture the regularities required for successful categorization. We report evidence that feature creation occurs in category learning and we describe the conditions that promote it. Feature creation can adapt flexibly to changing environmental demands and may be the origin of fixed feature repertoires. Implications for object categorization, conceptual development, chunking, constructive induction and formal models of dimensionality reduction are discussed. -------------------------------------------------------------- To help you decide whether you would be an appropriate commentator for this article, an electronic draft is retrievable from the World Wide Web or by anonymous ftp or gopher from the US or UK BBS Archive. Ftp instructions follow below. Please do not prepare a commentary on this draft. Just let us know, after having inspected it, what relevant expertise you feel you would bring to bear on what aspect of the article. The URLs you can use to get to the BBS Archive: http://www.princeton.edu/~harnad/bbs/ http://www.cogsci.soton.ac.uk/bbs/Archive/bbs.schyns.html ftp://ftp.princeton.edu/pub/harnad/BBS/bbs.schyns ftp://ftp.cogsci.soton.ac.uk/pub/bbs/Archive/bbs.schyns gopher://gopher.princeton.edu:70/11/.libraries/.pujournals To retrieve a file by ftp from an Internet site, type either: ftp ftp.princeton.edu or ftp 128.112.128.1 When you are asked for your login, type: anonymous Enter password as queried (your password is your actual userid: yourlogin at yourhost.whatever.whatever - be sure to include the "@") cd /pub/harnad/BBS To show the available files, type: ls Next, retrieve the file you want with (for example): get bbs.schyns When you have the file(s) you want, type: quit From becker at curie.psychology.mcmaster.ca Sun Feb 16 20:58:15 1997 From: becker at curie.psychology.mcmaster.ca (Sue Becker) Date: Sun, 16 Feb 1997 20:58:15 -0500 (EST) Subject: COMPUTATIONAL NEUROSCIENCE POSTDOCTORAL POSITION AVAILABLE Message-ID: COMPUTATIONAL NEUROSCIENCE POSTDOCTORAL POSITION AVAILABLE Department of Psychology McMaster University A postdoctoral position is open in the Psychology Department of McMaster University, Hamilton, Ontario. A multidisciplinary approach will be taken to develop biologically plausible models of hippocampal and neocortical memory systems. Projects will include developing simulations of cortical-cortical and subcortical-cortical interactions during learning and information storage. In addition to our focus on processing in hippocampal-cortical systems, we are also investigating and modelling the role of cortico-thalamic back-projections. Research projects will be conducted in close collaboration with R. Racine, a neuroscientist, S. Becker, a computational modeller, S. Haykin, an electrical engineer, and G. Gerstein, a computational neuroscientist. The primary goal of this collaborative effort is to build powerful learning algorithms in neural networks which are based on rules suggested by both memory research and physiology research (e.g. LTP work). Racine's laboratory has recently provided the first demonstrations of LTP in the neocortex and cortico-thalamic connections in chronic preparations. The rules that apply to LTP induction in these systems are quite different from those determined for the hippocampus. An optional component of this postdoctoral position would be participation in further experimental investigations of neocortical LTP in either slice or in vivo preparations. This project is funded by a collaborative research grant from the Natural Sciences and Engineering Research Council of Canada to R. Racine, G. Gerstein, S. Haykin and S. Becker. Please send curriculum vitae, expression of interest, and the names and e-mail or phone numbers of three references to Ron Racine at racine at mcmail.cis.mcmaster.ca From Gerhard.Paass at gmd.de Mon Feb 17 05:48:15 1997 From: Gerhard.Paass at gmd.de (Gerhard Paass) Date: Mon, 17 Feb 1997 11:48:15 +0100 Subject: CfP: Interdisciplinary College IK'97 Message-ID: <199702171048.AA02418@sein.gmd.de> Call for Participation Interdisciplinary College IK'97 This conference aims at establishing a link between four subjects: biology of neurons, neural information science, artificial intelligence and cognitive psychology. Leading scientists of each field provide introductory courses on each subject and give an in-depth account of current developments. In addition six interdisciplinary courses between each of the areas are offered, where two scientists of each field give an integrative view. Conference Language: German Time: March 15.-22.th 1997, Location: Guenne/Moehnesee, close to Dortmund/Germany Deadline for Registration: Feb. 28th (payment of conference fees) WWW: http://www.informatik.uni-ulm.de/ni/ik97.html From georg at ai.univie.ac.at Mon Feb 17 12:06:52 1997 From: georg at ai.univie.ac.at (Georg Dorffner) Date: Mon, 17 Feb 1997 18:06:52 +0100 (MET) Subject: Open position: NN for time series processing Message-ID: <199702171706.SAA21667@jedlesee.ai.univie.ac.at> The Neural Networks Group at the Austrian Research Institute for Artificial Intelligence in Vienna, Austria, has an immediate opening for a position on Neural Networks for Time Series Processing and Economic Modeling as part of the larger initiative "Adaptive Models in Economics and Management Science" (http://www.wu-wien.ac.at/am/am.html). The goal is to investigate the capabilities of recurrent neural nets in modeling economic processes, including their use in multi-agent simulations, with a focus on learning algorithms, adaptivity, and mixture-of-expert approaches. Required background: - knowledge in practical time series processing (e.g. ARMA, state space models, etc.) - basic knowledge in neural networks and programming - Ph.D. degree or comparable doctorate - research experience Desired background: - publications in the field - basic knowledge in economic modeling - basic knowledge in statistics - experience in a UNIX environment - basic understanding knowledge of German The position will be available for up to three years, with a monthly salary of about 18.000 ATS (approx. 1500 US$) per month, 14 times a year, after tax. Interested applicants should apply by email, mail or fax, no later than March 5, 1997 at the address below. Include a short vita, a list of publications, and any other information demonstrating your qualification. Send applications to: Georg Dorffner Austrian Research Institute for Artificial Intelligence Schottengasse 3 A-1010 Vienna phone: +43-1-53532810 fax: +43-1-5320652 email: georg at ai.univie.ac.at This position is funded by the Austrian Fonds zur Foerderung der wissenschaftlichen Forschung. From reggia at cs.umd.edu Mon Feb 17 15:11:29 1997 From: reggia at cs.umd.edu (James A. Reggia) Date: Mon, 17 Feb 1997 15:11:29 -0500 (EST) Subject: New Book: Neural Modeling of Brain and Cognitive Disorders Message-ID: <199702172011.PAA16491@avion.cs.umd.edu> NEURAL MODELING OF BRAIN AND COGNITIVE DISORDERS edited by James A Reggia (Univ. Maryland), Eytan Ruppin (Tel Aviv Univ.) & Rita Sloan Berndt (Univ. Maryland) During the last few years there has been a rapidly increasing interest in neural modeling of brain and cognitive disorders. This multidisciplinary book presents a variety of such models in neurology, neuropsychology and psychiatry. A review of work in this area is given first. Computational models are then presented of memory impairment in Alzheimer's disease, functional brain reorganization following a stroke, patterns of neural activity in epilepsy, disruption of language processes in aphasia and acquired dyslexia, altered cognitive processes in schizophrenia and depression, and related disorders. 425pp (approx.) 1996 981-02-2879-1 US$86 Ordering Information: Please e-mail or fax your order and VISA/MC/AMEX/Diners Club Card Number or alternatively enclose a cheque/bank draft and send to your nearest World Scientific Office: USA: World Scientific Publishing Co. Inc. 1060 Main Street, River Edge, NJ 07661, USA Toll-free: 1-800-227-7562 Toll-free fax: 1-888-977-2665 E-mail: wspc at wspc.com -UK: World Scientific Publishing Co. Ltd. 57 Shelton Street, Covent Garden, London WC2H 9HE, UK Tel: 44-171-836-0888 Fax: 44-171-836-2020 E-mail: sales at wspc2.demon.co.uk -Singapore: World Scientific Publishing Co. Ltd. Farrer Road P.O. Box 128, Singapore 912805 Tel: 65-382-5663 Fax: 65-382-5919 From zhangw at redwood.rt.cs.boeing.com Mon Feb 17 19:48:02 1997 From: zhangw at redwood.rt.cs.boeing.com (Wei Zhang) Date: Mon, 17 Feb 1997 16:48:02 -0800 Subject: Job: machine learning at Boeing Message-ID: <199702180048.QAA04330@rivendell.network-A> **Outstanding Machine Learning Researcher needed** The Boeing Company, the world's largest aerospace company, is actively working research projects in advanced computing technologies including projects involving NASA, FAA, Air Traffic Control, and Global Positioning as well as airplane and manufacturing research. The Research and Technology organization located in Bellevue, Washington, near Seattle, has an open position for a machine learning researcher. We are the primary computing research organization for Boeing and have contributed heavily to both short term technology advances and to long range planning and development. BACKGROUND REQUIRED: Machine Learning, Knowledge Discovery, Data Mining, Statistics, Artificial Intelligence or related field. RESEARCH AREAS: We are developing and applying techniques for data mining and statistical analyses of diverse types of data, including: safety incidents, flight data recorders, reliability, maintenance, manufacturing, and quality assurance data. These are not areas where most large R&D data mining efforts are currently focused. Research areas include data models, data mining algorithms, statistics, and visualization. Issues related to our projects also include pattern recognition, multidimensional time series, and temporal databases. We can achieve major practical impacts in the short-term both at Boeing and in the airline industry, which may result in a safer and more cost-effective air travel industry. A Ph.D. in Computer Science or equivalent experience is highly desirable for the position. We strongly encourage diversity in backgrounds including both academic and industrial experiences. Knowledge of machine learning, statistics, and data mining are important factors. Experience with databases and programming (C/C++, JAVA, and Splus) is desirable. APPLICATION: If you meet the requirements and you are interested, please send your resume via electronic e-mail in plain ASCII format to zhangw at redwood.rt.cs.boeing.com (Wei Zhang). You can also send it via US mail to Wei Zhang The Boeing Company PO Box 3707, MS 7L-66 Seattle, WA 98124-2207 Application deadline is April 30, 1997. The Boeing Company is an equal opportunity employer. From parodi at cptsu2.univ-mrs.fr Tue Feb 18 07:54:31 1997 From: parodi at cptsu2.univ-mrs.fr (Olivier Parodi) Date: Tue, 18 Feb 97 13:54:31 +0100 Subject: Cargese Summer School on Neural Information Processing Message-ID: <9702181254.AA19906@cptsu2.univ-mrs.fr> SUMMER SCHOOL on NEURONAL INFORMATION PROCESSING : FROM BIOLOGICAL DATA TO MODELLING AND APPLICATIONS Cargese -- Corse du Sud (France) June 30 -- July 12, 1997 organized at the INSTITUT D'ETUDES SCIENTIFIQUES DE CARGESE CNRS (UMS 820) Universite de Nice-Sophia Antipolis - Universite de Corte F 20130 CARGESE Sponsored by Centre National de la Recherche Scientifique DGA-DRET (French Ministry of Defence) Conseil Executif de la Corse THEME In the past ten years, statistical mechanics and dynamics of neuronal automata have been extensively studied. Most of the work has been based on over-simplified models of neurons. Recent developments in Neurosciences have, however, considerably modified our knowledge of both the operating modes of neurons and information processing in the cortex. Multi-unit recordings have allowed precise temporal correlations to be detected, within temporal windows of the order of 1 ms. Simultaneously, oscillations corresponding to a quasi-periodic spike-firing, synchronized over several visual cortical areas, have been observed with anaesthesied cats and with monkeys. Last but not least, recent work on the neuronal operating modes have emphasized the role played by the dendritic arborization. These developments have led to considerable interest for coding scheme relying on precise spatio-temporal patterns both from the theoretical and experimental points of view. This prompts us to consider, for information processing, new models which would proceed, e.g., from a synchronous detection of correlated spike firing, and could be particularly robust against noise. Such models might bring about original technical applications for information processing and control. Further developments in this field may be of major importance for our understanding of the basic mechanisms of perception and cognition. They should also lead to new concepts in applications directed towards artificial perception and pattern recognition. Up to now, artificial systems for pattern recognition are far from reaching the standards of human vision. Systems based on a temporal coding by spikes may now be expected to bring about major improvements in this field. The aim of the school is to provide students and people engaged in both applied and basic research with state of the art in every relevant field (Neurosciences, Physics, Mathematics, Information and Control Theory) and to encourage further interdisciplinary and international exchanges. PRELIMINARY PROGRAM Pr. M. Abeles (Jerusalem) - Spike Correlations and synfire chains, Pr. P. Combe (Marseille) - Learning: a geometrical approach, Pr. J. Demongeot (Grenoble) - Dynamics of modular architectures, Pr. L. van Hemmen (Munchen) - Coding by spikes, Pr. J. Herault (Grenoble) - Information processing in the retina, Pr. J. Hertz (Copenhague) - Temporal coding and learning, Pr. A. Holley (Lyon) - Information processing in the olfactory mammalian system, Pr. C. von der Malsburg (Bochum) - Binding and Synchrony, Dr. C. Masson (Paris) - Cellular mechanism of odor discrimination, Dr. J. P. Nadal (Paris) - Information theory and neuronal architecture, Dr. O. Parodi (Marseille) - Temporal coding and correlation detection, Pr. S. Solla (ATT) - Dynamics of on-line learning processes. OFFICIAL LANGUAGES The official languages of the School are English and French. Lectures will given in English. DIRECTOR OF THE SCHOOL Olivier PARODI, Centre de Physique Theorique CNRS-Luminy, case 907, F 13288 MARSEILLE CEDEX 09, France Tel. (33) 4 91 26 95 30 Fax (33) 4 91 26 95 53 e-mail parodi at cpt.univ-mrs.fr REGISTRATION FEES Students: free CNRS and members of CNRS institutes: free University: 1500 FF Industry: 2500 FF ACCOMMODATION GRANTS 1 - The School is sponsored by the Formation Permanente du CNRS, which can support accomodation expenses of at least 16 CNRS participants. 2 - The Organizing Committee will consider grants for students and young participants. PRACTICAL INFORMATION The school will be held at the Institut d'Etudes Scientifiques de Cargese. Lectures and Seminars will be given from 9 am to 12:30 pm and from 4 to 7:15 pm.(except on Sunday) from July 1 to 11 included. All participants are expected to arrive on June 30 and leave on July 12. TRAVEL : Cargese is located approximatively 60 kms North of Ajaccio. The best way to get to Cargese is to reach Ajaccio. You are asked to make your own travelling arrangements. However, in order to reduce the cost of your travel, two groups will be organized on regular flights between Paris and Ajaccio and Marseille and Ajaccio on June 30 and July 12 (approx. 1200-1400 FF for a Paris- Ajaccio return ticket). We also consider renting buses for the Ajaccio-Cargese and return journeys. Details will be sent later. ACCOMODATION : The Institute is located 2 km south of the village, on the sea shore. On working days, lunches will be served at the Institute (800 FF for the session, including refreshment and coffe breaks). There are various housing possibilities: - shared room at the Institute or in apartments in the village - 1920 FF (per person for the session) - single room in shared apartments in the village - 2640 FF for the session - rented apartment in the village for your family - 270 to 500 FF per day - hotels in the village - 250 to 400 FF per person per day - camping on the grounds of the Institute - 20 FF per person per day. You have to bring your own equipment, showers are at your disposal. Breakfast will be served at the Institute for people staying on the grounds. WARNING: There are no banks nor cash machines in Cargese and credit cards are not accepted everywhere. Think to bring enough French cash before leaving Paris, Marseille or Ajaccio. NOTE : We cannot provide accomodation before or after the dates of the School. POSTER SESSION : One or several poster sessions will be organized. Participants are encouraged to prepare posters on their own work. REGISTRATION : You have to fill the enclosed application and return it by e-mail to cargese at cpt.univ-mrs.fr BEFORE MARCH 31, 1997 with, if necessary, a letter justifying your grant request. The School will accept up to 50 students, including those supported by the Formation Permanente of CNRS. In case of over-demand, participants will be selected by the Scientific Committee, with a balance between junior and senior students, and a preference for students carrying an active research in the field. ------------------ http://cpt.univ-mrs.fr/Cargese ---------------------- ***************************************************************************** ***************************************************************************** French Version Version Francaise (TEX) a compiler en "plain TEX" sur une station UNIX, un MAC avec Textures ou un PC. ***************************************************************************** \nopagenumbers \vsize 26.5 true cm \hsize 18.5 true cm \hoffset -1.2 true cm \voffset -1 cm \null {\bf \centerline{ECOLE D'ETE} \vskip 1 true cm \centerline{TRAITEMENT NEURONAL DE L'INFORMATION} \medskip \centerline{Des donn\'ees biologiques \`a la mod\'elisation et aux applications} \vskip 1 true cm \centerline{Carg\`ese, Corse du Sud (France)} \par \centerline{30 juin -- 12 juillet 1997} \bigskip \centerline{organis\'ee \`a} \bigskip \centerline{L'INSTITUT D'ETUDES SCIENTIFIQUES DE CARGESE} } \smallskip \centerline{CNRS (UMS 820) -- Universit\'e de Nice-Sophia Antipolis -- Universit\'e de Corte} \smallskip \centerline{F 20130 Carg\`ese} \vskip 1 true cm \noindent Subventionn\'ee par \par Le Centre National de la Recherche Scientifique \par La Direction Scientifique de la DGA-DRET (Minist\`ere de la D\'efense) \par La Collectivit\'e Territoriale Corse \vskip 1 true cm\noindent {\bf THEME} \medskip Depuis une dizaine d'ann\'ees, de nombreuses \'etudes ont port\'e sur la m\'ecanique statistique et la dynamique d'au\-to\-ma\-tes neuronaux, bas\'ees le plus souvent sur une mod\'elisation relativement ancienne du neurone. Mais les d\'eveloppements r\'ecents en Neurosciences ont consid\'erablement modifi\'e nos connaissances, \`a la fois sur le neurone et sur le traitement de l'information dans le cortex. \par Les techniques d'enregistrement multi-\'electrodes ont permis de d\'etecter des corr\'elations temporelles tr\`es pr\'ecises dans des fen\^etres de l'ordre de la milliseconde. En m\^eme temps des {\it oscillations}, correspondant \`a des \'emissions quasi-p\'eriodiques de potentiels d'action synchronis\'ees \`a travers tout le cortex visuel, ont \'et\'e observ\'ees chez des chats anesth\'esi\'es et chez des singes. Parall\`element, des travaux portant sur le fonctionnement du neurone ont soulign\'e le r\^ole important jou\'e par l'arborisation dendritique, r\^ole compl\`etement pass\'e sous silence dans les mod\`eles classiques d'automates neuronaux. \par Ces d\'eveloppements ont donn\'e un essor consid\'erable \`a des \'etudes, tant th\'eoriques qu'exp\'erimentales, faisant intervenir un codage par des configurations spatio-temporelles pr\'ecises. On est ainsi amen\'e \`a envisager de nouveaux mod\`eles de traitement de l'information, bas\'es, par exemple, sur la d\'etection synchrone de signaux corr\'el\'es, et qui pourraient \^etre particuli\`erement robustes vis-\`a-vis du bruit. Ces mod\`eles pourraient conduire \`a des applications technologiques originales en mati\`ere de traitement de l'information et de contr\^ole. \par Le but de cette \'ecole est d'apporter aux \'etudiants et aux chercheurs, tant fondamentaux qu'ap\-pli\-qu\'es, un enseignement de pointe couvrant, dans ce domaine, l'ensemble des disciplines concern\'ees (Neurosciences, Physique, Math\'ematiques, Th\'eories de l'Information et du Contr\^ole), et de favoriser des coop\'erations interdisciplinaires et internationales. \vskip .8 true cm\noindent {\bf PROGRAMME PRELIMINAIRE} \medskip \par Pr. M. Abeles (Jerusalem) -- Spike Correlations and synfire chains, \par Pr. P. Combe (Marseille) -- Learning: a geometrical approach, \par Pr. J. Demongeot (Grenoble) -- Dynamics of modular architectures, \par Pr. L. van Hemmen (Munchen) -- Coding by spikes, \par Pr. J. H\'erault (Grenoble) -- Information processing in retina, \par Pr. J. Hertz (Copenhague) -- Temporal coding and learning, \par Pr. A. Holley (Lyon) -- Information processing in the olfactory mammalian system, \par Pr. C. von der Malsburg (Bochum) -- Binding and Synchrony, \par Dr. C. Masson (Paris) -- Cellular mechanism of odor discrimination, \par Dr. J. P. Nadal (Paris) -- Information theory and neuronal architecture, \par Dr. O. Parodi (Marseille) -- Temporal coding and correlation detection, \par Pr. S. Solla (ATT) -- Dynamics of on-line learning processes \vfill \eject \noindent {\bf LANGUES OFFICIELLES} \smallskip Les langues officielles de l'Ecole sont l`anglais et le fran\c cais. Les cours seront donn\'es en anglais \bigskip\noindent {\bf DIRECTEUR DE L'ECOLE} \smallskip Olivier PARODI, Centre de Physique Th\'eorique \par CNRS-Luminy, case 907, F 13288 MARSEILLE CEDEX 09, France \par Tel. (33) 4 91 26 95 30 Fax (33) 4 91 26 95 53 \par e-mail parodi at cpt.univ-mrs.fr \bigskip\noindent {\bf DROITS D'INSCRIPTION} \smallskip \par Etudiants: n\'eant \par Personnels CNRS et membres de laboratoires soutenus par le CNRS: n\'eant \par Autres universitaires: 1500 FF \par Secteur industriel: 2500 FF \bigskip\noindent {\bf ALLOCATIONS} \smallskip \item{1 -} L'Ecole est subventionn\'ee par la Formation permanente du CNRS. Les frais de s\'ejour d'au moins 16 participants CNRS -- ou, \`a d\'efaut, de participants en provenance de Laboratoires soutenus par le CNRS -- pourront \^etre pris en charge par la Formation Permanente de la XIIe D\'el\'egation. Les frais de transport des personnels CNRS seront alors pris en charge par la Formation Permanente de leur D\'el\'egation. \item{2 -} Le Comit\'e d'Organisation envisage d'attribuer des bourses aux jeunes chercheurs. \bigskip\noindent {\bf INFORMATIONS PRATIQUES} \smallskip L'Ecole est organis\'ee \`a l'Institut d'Etudes Scientifiques de Carg\`ese. Les Cours et les S\'eminaires auront lieu de 9h \`a 12h30 et de 16h \`a 19h15 tous les jours, sauf le Dimanche, du 1er juillet au 11 juillet inclus. Tous les participants sont attendus le 30 juin et leur d\'epart est pr\'evu le 12 juillet. \medskip\noindent VOYAGE: \par Carg\`ese est situ\'e \`a environ 60 km au Nord d'Ajaccio. La meilleure mani\`ere de s'y rendre est de gagner d'abord Ajaccio. Vous \^etes responsable de l'organisation de votre voyage, aller et retour. Cependant, afin d'en r\'eduire le co\^ut, des vols de groupe sur ligne r\'eguli\`ere seront organis\'es entre Paris et Ajaccio et Marseille et Ajaccio les 30 juin et 12 juillet (Prix approximatif 1200 \`a 1400 FF pour un aller et retour Paris-Ajaccio). Nous envisageons aussi de louer des cars pour les aller et retour Ajaccio-Carg\`ese. Les d\'etails seront communiqu\'es ult\'erieurement. \medskip\noindent LOGEMENT: \par L'Institut est situ\'e \`a 2 km du village au bord de la mer. Les jours de cours, le d\'ejeuner est servi sur place (800 FF pour la session y compris les pauses-caf\'e). Les possibilit\'es d'h\'ebergement sont diverses: \medskip \item{- } Chambre partag\'ee avec un coll\`egue \`a l'Institut ou dans un appartement au village (1920 FF pour la dur\'ee de la session), \item{- } Chambre individuelle dans un appartement partag\'e (2640 FF pour la session), \item{- } Appartement au village en location pour votre famille (270 \`a 500 FF par jour) \item{- } H\^otel dans le village (250 \`a 400 FF par jour et par personne) \item{- } Camping sur le terrain m\^eme de l'Institut, avec votre mat\'eriel individuel (20 FF par jour). \medskip\noindent {\it ATTENTION: Pas de banque ni de distributeur automatique de billet \`a Carg\`ese, et les cartes de cr\'edit ne sont pas toujours accept\'ees. Pensez \`a vous munir d'argent liquide.} \medskip\noindent NB: Nous n'organisons aucun h\'ebergement en dehors des dates de l'Ecole. \bigskip\noindent {\bf COMMUNICATIONS AFFICH\'EES} \par Une ou plusieurs s\'eances de communications par affiches seront organis\'ees. Les participants sont invit\'es \`a pr\'eparer des affiches sur leur propre travail de recherche. \bigskip\noindent {\bf INSCRIPTION} \par Le formulaire ci-joint doit \^etre renvoy\'e au plus tard le {\bf 31 mars 1997} par courrier \'electronique \`a \break {\bf cargese at cpt.univ-mrs.fr}, accompagn\'e, \'eventuellement, d'une lettre justifiant votre demande de bourse ou de prise en charge par la Formation Permanente du CNRS. \par Le nombre de participants \`a l'Ecole, enseignants non compris, est limit\'e \`a cinquante, nombre incluant les stagiaires subventionn\'es par la Formation Permanente du CNRS. La s\'election des candidats sera effectu\'ee par le Comit\'e Scientifique, qui veillera \`a l'\'equilibre entre jeunes chercheurs et chercheurs confirm\'es, et donnera une priorit\'e aux participants engag\'es dans une recherche active dans le domaine. \medskip \noindent {\bf WEB} \par Toutes les informations concernant l'Ecole sont disponibles \`a l'URL \par {\bf http://cpt.univ-mrs.fr/Cargese/} \end *************************************************************************** *************************************************************************** ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||| REGISTRATION FORM BULLETIN D'INSCRIPTION NEURONAL INFORMATION PROCESSING TRAITEMENT NEURONAL DE L'INFORMATION NAME / NOM First name / Prenom (Ms/Mr) Nationality/Nationalite date of birth / date INSTITUTE ADDRESS / LABORATOIRE Tel. Fax e-mail HOME ADDRESS / Adresse personnelle Tel. Next of kin / Personne a prevenir en cas d'accident : DIPLOMA /Niveau Universitaire PRESENT POSITION / EMPLOI ACTUEL ? PRESENT RESEARCH / RECHERCHES ACTUELLES MAIN PUBLICATIONS / Publications Principales (up to five recent publications) Will your Institution pay for / Votre Laboratoire subventionne-t-il: your travel (Yes/No) ? votre voyage (Oui/Non) ? your accommodation expenses (Yes/No) ? vos frais de sejour ? (Oui/Non) ? Only for members of the CNRS / Pour les participants CNRS, ou relevant d'un laboratoire soutenu par le CNRS, demandez-vous la prise en charge de vos frais de sejour et de transport par la Formation Permanente du CNRS ? For Junior applicants, do you request a grant ? (Yes/No) ? Pour les jeunes chercheurs, etes-vous candidat a une bourse ? (Oui/Non) RECOMMENDATION : Junior applicants should arrange for a letter of recommendation from a senior scientist who had the oppotunity to judge their work. This letter should be sent by e-mail to cargese at cpt.univ-mrs.fr. RECOMMANDATION: Les jeunes chercheurs doivent fournir une lettre de recommandation d'un chercheur confirme. Cette lettre doit etre envoyee par courrier electronique a cargese at cpt.univ-mrs.fr YOU ARE RECOMMENDED BY / Vous etes recommande par ? TRAVEL/VOYAGE: Will you take a group flight / Prendrez-vous un vol de groupe ? Paris-Ajaccio-Paris : Number/Nombre ? Marseille-Ajaccio-Marseille : Number/Nombre ? HOUSING (Specify your choice): LOGEMENT (Precisez votre choix): Pour les participants francophones: Etes-vous capable de suivre un cours en anglais ? De participer a une discussion en anglais ? DEADLINE / DATE LIMITE D'INSCRIPTION : THIS APPLICATION HAS TO BE SENT BY E-MAIL NOT LATER THAN MARCH 31, 1997 TO cargese at cpt.univ-mrs.fr CE FORMULAIRE DOIT ETRE RENVOYE PAR COURRIER ELECTRONIQUE AU PLUS TARD LE 31 MARS 1997 A cargese at cpt.univ-mrs.fr You will receive an answer by e-mail before April 30. Vous recevrez une reponse avant le 30 avril. e-mail contact for special requests only / contact par courrier electronique pour des problemes particuliers: cargese at cpt.univ-mrs.fr From ruderman at salk.edu Tue Feb 18 14:38:21 1997 From: ruderman at salk.edu (Dan Ruderman) Date: Tue, 18 Feb 1997 11:38:21 -0800 Subject: Preprint available Message-ID: <330A052D.745F682F@salk.edu> The following preprint is available: Origins of scaling in natural images Daniel L. Ruderman One of the most robust qualities of our visual world is the scale-invariance of natural images. Not only has scaling been found in different visual environments, but the phenomenon also appears to be calibration independent. This paper proposes a simple property of natural images which explains this robustness: They are collages of regions corresponding to statistically independent ``objects''. Evidence is provided for these objects having a power-law distribution of sizes within images, from which follows scaling in natural images. It is commonly suggested that scaling instead results from edges, each with power spectrum 1/k^2. This hypothesis is refuted by example. (To appear in Vision Research) It may be found at my website: http://quake.usc.edu/~dlr/papers.html -- Dr. Dan Ruderman ruderman at sloan.salk.edu The Salk Institute (VCL) 10010 N. Torrey Pines Road La Jolla, CA 92037 (619) 453-4100 x2025 FAX: (619) 455-7933 From pelillo at dsi.unive.it Wed Feb 19 03:33:05 1997 From: pelillo at dsi.unive.it (Marcello Pelillo) Date: Wed, 19 Feb 1997 09:33:05 +0100 (MET) Subject: EMMCVPR'97 (Venice) - Program Message-ID: <199702190833.JAA16875@oink.dsi.unive.it> A non-text attachment was scrubbed... Name: not available Type: text Size: 8976 bytes Desc: not available Url : https://mailman.srv.cs.cmu.edu/mailman/private/connectionists/attachments/00000000/b1f43ccc/attachment.ksh From gareth at stat.stanford.edu Wed Feb 19 17:25:57 1997 From: gareth at stat.stanford.edu (Gareth James) Date: Wed, 19 Feb 1997 14:25:57 -0800 (PST) Subject: Paper on Generalized Bias/Variance Decomposition Message-ID: The following paper is available in postscript format at http://stat.stanford.edu/~gareth under my papers section. Generalizations of the Bias/Variance Decomposition for Prediction Error Gareth James and Trevor Hastie (Stanford University) The bias and variance of a real valued random variable, using squared error loss, are well understood. However because of recent developments in classification techniques it has become desirable to extend these concepts to general random variables and loss functions. The 0-1 (misclassification) loss function with categorical random variables has been of particular interest. We explore the concepts of variance and bias and develop a decomposition of the prediction error into functions of the systematic and variable parts of our predictor. After providing some examples we conclude with a discussion of the various definitions that have been proposed. From carmesin at schoner.physik.uni-bremen.de Thu Feb 20 11:09:05 1997 From: carmesin at schoner.physik.uni-bremen.de (Hans-Otto Carmesin) Date: Thu, 20 Feb 1997 17:09:05 +0100 Subject: Theory and Experiment papers Message-ID: <199702201609.RAA03423@schoner.physik.uni-bremen.de> Two corresponding experimental and theoretical papers concerning -- Continuous Phase Transitions in Multistable Perception: Neural Network and Experiment -- are now available. Based on a neural network theory, we predicted a continuous phase transition for multistable perception (H.-O. Carmesin: Theorie neuronaler Adaption, Koester, Berlin 1994,1996). Then we discovered this continuous transition experimentally, observed additional dynamical effects at the sensitive continuous transition and improved the neural network model accordingly. The full neural network theory is presented in the paper - A neural network model for stroboscopic alternative motion - by Hans-Otto Carmesin and Stefan Arndt in Biological Cybernetics 75, 239-251 (1996). ABSTRACT: A neural network which models multistable perception is presented. The network consists of sensor and inner neurons. The dynamics is established by a stochastic neuronal dynamics, a formal Hebb-type coupling dynamics and a resource mechanism that corresponds to saturation effects in perception. >From this a system of coupled differential equations is derived and analyzed. Single stimuli are bound to exactly one percept, even in ambiguous situations where multistability occurs. The network exhibits discontinuous as well as continuous phase transitions and models various empirical findings, including the percepts of succession, alternative motion and simultaneity; the percept of oscillation is explained by oscillating percepts at a continuous phase transition. The experimental methods and results are presented in the paper - Continuous phase transitions in the perception of multistable visual patterns - by Peter Kruse, Hans-Otto Carmesin, Lars Pahlke, Daniel Strber and Michael Stadler in Biological Cybernetics 75, 321-330 (1996). ABSTRACT: The phenomenon of alternative stroboscopic motion exhibits five different percepts that are seen with an increase in frequency of presentation: (a) succession, (b) fluttering motion, (c) reversible clockwise and counterclockwise motion, (d) oppositional motion and (e) simultaneity. From a synergetic point of view the increase in frequency is a control parameter and different percepts are order parameters with phase transitions in between. The neural network theory of Carmesin and Arndt is applied to receive predictions about hysteresis and phase transitions between these order parameters. Empirical data show the different motion percepts (b), (c) and (e) have lognormal distributions. Following the theoretical model, it is argued that there are three different phases, (a), (c) and (e), with two continuous phase transitions, (b) and (d), between them. The experimental data substantially match the theoretical assumptions. Some free reprints are available and can be requested via email from carmesin at theo.physik.uni-bremen.de. Moreover, two related papers are available online in the www at http://schoner.physik.uni-bremen.de/~carmesin/docs/zkw9505.ps and http://schoner.physik.uni-bremen.de/~carmesin/docs/cyber2u.ps. Hans-Otto Carmesin Institute for Theoretical Physics, University Bremen, 28334 Bremen, Germany, WWW: http://schoner.physik.uni-bremen.de/~carmesin/ From prefenes at lbs.ac.uk Fri Feb 21 19:55:04 1997 From: prefenes at lbs.ac.uk (Paul Refenes) Date: Fri, 21 Feb 1997 19:55:04 BST Subject: Preprints available - Neural Networks in the Capital Markets Message-ID: <5E6F0DD62D7@neptune.lbs.ac.uk> Neural Networks in the Capital Markets The following NNCM-96 pre-prints are now available on request. Please send your postal address to: boguntula at lbs.lon.ac.uk =================================================== NEURAL MODEL IDENTIFICATION, VARIABLE SELECTION AND MODEL ADEQUACY A-P. N. REFENES, A. D. ZAPRANIS AND J. UTANS Department of Decision Science London Business School Regents Park, London, NW1 4SA, UK In recent years an impressive array of publications have appeared claiming considerable successes of neural networks in modeling financial data but skeptical practitioners and statisticians are still raising the question of whether neural networks really are "a major breakthrough or just a passing fad". A major reason for this is the lack of procedures for performing tests for mispecified models, and tests of statistical significance for the various parameters that have been estimated, which makes it difficult to assess the model's significance and the possibility that any short term successes that are reported might be due to "data mining". In this paper we describe a methodology for neural model identification which facilitates hypothesis testing at two levels: model adequacy and variable significance. The methodology includes a model selection procedure to produce consistent estimators, a variable selection procedure based on variable significance testing and a model adequacy procedure based on residuals analysis. ================================================ SPECIFICATION TESTS FOR NEURAL NETWORKS: A CASE STUDY IN TACTICALASSET ALLOCATION A. D. ZAPRANIS, J. UTANS, A-P. N. REFENES Department of Decision Science London Business School Regents Park, London, NW1 4SA, UK A case study in tactical asset allocation is used to introduce a methodology for neural model identification including model specification and variable selection. Neural models are contrasted to multiple linear regression on the basis of model identification. The results indicate the presence of non-linear relationships between the economic variables and asset class returns. ============================================= From harnad at cogsci.soton.ac.uk Sun Feb 23 12:11:28 1997 From: harnad at cogsci.soton.ac.uk (Stevan Harnad) Date: Sun, 23 Feb 97 17:11:28 GMT Subject: Psycoloquy Call for Papers Message-ID: <18205.9702231711@cogsci.ecs.soton.ac.uk> PSYCOLOQUY CALL FOR PAPERS PSYCOLOQUY is a refereed electronic journal (ISSN 1055-0143) sponsored on an experimental basis by the American Psychological Association and currently estimated to reach a readership of 50,000. PSYCOLOQUY publishes reports of new ideas and findings on which the author wishes to solicit rapid peer feedback, international and interdisciplinary ("Scholarly Skywriting"), in all areas of psychology and its related fields (biobehavioral science, cognitive science, neuroscience, social science, etc.). All contributions are refereed. Target article length should normally not exceed 500 lines [c. 4500 words]. Commentaries and responses should not exceed 200 lines [c. 1800 words]. All target articles, commentaries and responses must have (1) a short abstract (up to 100 words for target articles, shorter for commentaries and responses), (2) an indexable title, (3) the authors' full name(s), institutional address(es) and URL(s). In addition, for target articles only: (4) 6-8 indexable keywords, (5) a separate statement of the authors' rationale for soliciting commentary (e.g., why would commentary be useful and of interest to the field? what kind of commentary do you expect to elicit?) and (6) a list of potential commentators (with their email addresses). All paragraphs should be numbered in articles, commentaries and responses (see format of already published articles in the PSYCOLOQUY archive; line length should be < 80 characters, no hyphenation). It is strongly recommended that all figures be designed so as to be screen-readable ascii. If this is not possible, the provisional solution is the less desirable hybrid one of submitting them as .gif .jpeg .tiff or postscript files (or in some other universally available format) to be printed out locally by readers to supplement the screen-readable text of the article. PSYCOLOQUY also publishes multiple reviews of books in any of the above fields; these should normally be the same length as commentaries, but longer reviews will be considered as well. Book authors should submit a 500-line self-contained Precis of their book, in the format of a target article; if accepted, this will be published in PSYCOLOQUY together with a formal Call for Reviews (of the book, not the Precis). The author's publisher must agree in advance to furnish review copies to the reviewers selected. Authors of accepted manuscripts assign to PSYCOLOQUY the right to publish and distribute their text electronically and to archive and make it permanently retrievable electronically, but they retain the copyright, and after it has appeared in PSYCOLOQUY authors may republish their text in any way they wish -- electronic or print -- as long as they clearly acknowledge PSYCOLOQUY as its original locus of publication. However, except in very special cases, agreed upon in advance, contributions that have already been published or are being considered for publication elsewhere are not eligible to be considered for publication in PSYCOLOQUY, Please submit all material to psyc at pucc.princeton.edu http://www.princeton.edu/~harnad/psyc.html http://cogsci.soton.ac.uk/psyc ftp://ftp.princeton.edu/pub/harnad/Psycoloquy ftp://cogsci.soton.ac.uk/pub/harnad/Psycoloquy gopher://gopher.princeton.edu/11/.libraries/.pujournals news:sci.psychology.journals.psycoloquy ---------------------------------------------------------------------- CRITERIA FOR ACCEPTANCE: To be eligible for publication, a PSYCOLOQUY target article should not only have sufficient conceptual rigor, empirical grounding, and clarity of style, but should also offer a clear rationale for soliciting Commentary. That rationale should be provided in the author's covering letter, together with a list of suggested commentators. A target article can be (i) the report and discussion of empirical research; (ii) an theoretical article that formally models or systematizes a body of research; or (iii) a novel interpretation, synthesis, or critique of existing experimental or theoretical work. Rrticles dealing with social or philosophical aspects of the behavioral and brain sciences are also eligible.. The service of Open Peer Commentary will be primarily devoted to original unpublished manuscripts. However, a recently published book whose contents meet the standards outlined above may also be eligible for Commentary. In such a Multiple Book Review, a comprehensive, 500-line precis by the author is published in advance of the commentaries and the author's response. In rare special cases, Commentary will also be extended to a position paper or an already published article dealing with particularly influential or controversial research. Submission of an article implies that it has not been published or is not being considered for publication elsewhere. Multiple book reviews and previously published articles appear by invitation only. The Associateship and professional readership of PSYCOLOQUY are encouraged to nominate current topics and authors for Commentary. In all the categories described, the decisive consideration for eligibility will be the desirability of Commentary for the submitted material. Controversially simpliciter is not a sufficient criterion for soliciting Commentary: a paper may be controversial simply because it is wrong or weak. Nor is the mere presence of interdisciplinary aspects sufficient: general cybernetic and "organismic" disquisitions are not appropriate for PSYCOLOQUY. Some appropriate rationales for seeking Open Peer Commentary would be that: (1) the material bears in a significant way on some current controversial issues in behavioral and brain sciences; (2) its findings substantively contradict some well-established aspects of current research and theory; (3) it criticizes the findings, practices, or principles of an accepted or influential line of work; (4) it unifies a substantial amount of disparate research; (5) it has important cross-disciplinary ramifications; (6) it introduces an innovative methodology or formalism for consideration by proponents of the established forms; (7) it meaningfully integrates a body of brain and behavioral data; (8) it places a hitherto dissociated area of research into an evolutionary or ecological perspective; etc. In order to assure communication with potential commentators (and readers) from other PSYCOLOQUY specialty areas, all technical terminology must be clearly defined or simplified, and specialized concepts must be fully described. NOTE TO COMMENTATORS: The purpose of the Open Peer Commentary service is to provide a concentrated constructive interaction between author and commentators on a topic judged to be of broad significance to the biobehavioral science community. Commentators should provide substantive criticism, interpretation, and elaboration as well as any pertinent complementary or supplementary material, such as illustrations; all original data will be refereed in order to assure the archival validity of PSYCOLOQUY commentaries. Commentaries and articles should be free of hyperbole and remarks ad hominem. STYLE AND FORMAT FOR ARTICLES AND COMMENTARIES TARGET ARTICLES: should not exceed 500 lines (~4500 words); commentaries should not exceed 200 lines (1800 words), including references. Spelling, capitalization, and punctuation should be consistent within each article and commentary and should follow the style recommended in the latest edition of A Manual of Style, The University of Chicago Press. It may be helpful to examine a recent issue of PSYCOLOQUY. All submissions must include an indexable title, followed by the authors' names in the form preferred for publication, full institutional addresses and electronic mail addresses, a 100-word abstract, and 6-12 keywords. Tables and diagrams should be made screen-readable wherever possible (if unavoidable, printable postscript files may contain the graphics separately). All paragraphs should be numbered, consecutively. No line should exceed 72 characters, and a blank line should separate paragraphs. REFERENCES: Bibliographic citations in the text must include the author's last name and the date of publication and may include page references. Complete bibliographic information for each citation should be included in the list of references. Examples of correct style are: Brown(1973); (Brown 1973); Brown 1973; 1978); (Brown 1973; Jones 1976); (Brown & Jones 1978); (Brown et al. 1978). References should be typed on a separate sheet in alphabetical order in the style of the following examples. Do not abbreviate journal titles. Kupfermann, I. & Weiss, K. (1978) The command neuron concept. Behavioral and Brain Sciences 1:3-39. Dunn, J. (1976) How far do early differences in mother-child relations affect later developments? In: Growing point in ethology, ed. P. P. G. Bateson & R. A. Hinde, Cambridge University Press. Bateson, P. P. G. & Hinde, R. A., eds. (1978) Growing points in ethology, Cambridge University Press. EDITING: PSYCOLOQUY reserves the right to edit and proof all articles and commentaries accepted for publication. Authors of articles will be given the opportunity to review the copy-edited draft. Commentators will be asked to review copy-editing only when changes have been substantial. --------------------------------------------------------------------------- Prof. Stevan Harnad psyc at pucc.princeton.edu Editor, Psycoloquy phone: +44 1703 594-583 fax: +44 1703 593-281 Department of Psychology http://cogsci.soton.ac.uk/psyc University of Southampton http://www.princeton.edu/~harnad/psyc.html Highfield, Southampton ftp://ftp.princeton.edu/pub/harnad/Psycoloquy SO17 1BJ UNITED KINGDOM ftp://cogsci.soton.ac.uk/pub/harnad/Psycoloquy news:sci.psychology.journals.psycoloquy gopher://gopher.princeton.edu/11/.libraries/.pujournals Sponsored by the American Psychological Association (APA) From payman at u.washington.edu Mon Feb 24 20:35:45 1997 From: payman at u.washington.edu (Payman Arabshahi) Date: Mon, 24 Feb 1997 17:35:45 -0800 (PST) Subject: CIFEr'97 Tutorials - New York, March 23, 1997 Message-ID: <199702250135.RAA01799@saul7.u.washington.edu> Computational Intelligence in Financial Engineering Conference CIFEr'97 March 23-25, 1997 Crowne Plaza Manhattan, New York City http://www.ieee.org/nnc/cifer97 Registration information: Barbara Klemm CIFEr'97 Secretariat Meeting Management 2603 Main Street, Suite # 690 Irvine, California 92714 Tel: (714) 752-8205 or (800) 321-6338 Fax: (714) 752-7444 Email: Meetingmgt at aol.com TUTORIALS ---------------------------------------------------------------------------- Risk Management Jan W. Dash, Ph.D. Director Quantitative Analysis Global Risk Management Smith Barney This tutorial will cover 1) characterization of risks in finance: market risk (interest rates, FX rates, equity indices, spreads), trading risk, systems risk (software, hardware, vendors), model risk, and 2) quantitative measurement of risk: the Greeks (Delta, Gamma, Vega), the partial Greeks (Ladders), the new Greeks (Exotics), dollars at risk (n-Sigma analysis), correlations, static scenario analysis, dynamic scenario analysis, Monte Carlo risk analysis, beginnings of risk standards, DPG, Risk Metrics, and 3) case study of risk: the Viacom CVR Options and 4) pricing and hedging for interest rate derivatives. ---------------------------------------------------------------------------- An Introduction to OTC Derivatives and Their Applications John F. Marshall, Ph.D. Executive Director International Association of Financial Engineers This tutorial is for persons with little prior exposure to derivative instruments. It will focus on the basic products, how they trade, and how they are used. It will be largely non-quantitative. The tutorial will examine how derivatives are used by financial engineers for risk management purposes, investment purposes, cash flow management, and creating structured securities. The use of derivatives to circumvent market imperfections, such as asymmetric taxes and transaction costs, will also be demonstrated. The primary emphasis of the tutorial will be swaps (including interest rate swaps, currency swaps, commodity swaps, equity swaps, and macroeconomic swaps). Applications of OTC options, including caps and floors and digital options will also be examined, but to a lesser extent. ---------------------------------------------------------------------------- GARCH Modeling of Financial Time Series R. Douglas Martin, Ph.D. Professor of Statistics, University of Washington Chief Scientist, Data Analysis Products Division of MathSoft, Inc. This tutorial provides an introduction to univariate and multivariate generalized autoregressive heteroscedastic (GARCH) modeling of financial returns time series data, with a focus on modeling conditional volatilities and correlations. Basic aspects of the various models are discussed, including: conditions for stationarity, optimization techniques for maximum likelihood estimation of the models, use of the estimated conditional standard deviations for value-at-risk calculations and options pricing, use of conditional correlations in obtaining conditional volatilities for portfolios. Examples are provided using the S+GARCH object-oriented toolkit for GARCH modeling. ---------------------------------------------------------------------------- Time Series Tools for Finance Andreas Wiegend, Ph.D. Professor, Stern School of Business, New York University This tutorial presents a unifying view of the recent advances of neuro-fuzzy, and other machine learning techniques for time series and finance. It is given jointly by Prof. Andreas Wiegend (Stern School of Business, NYU), and Dr. Georg Zimmerman (Siemens AG, Munich), and presents both conceptual aspects of time series modeling, specific tricks for financial engineering problems, and software engineering aspects for building a trading system. ---------------------------------------------------------------------------- An Introduction to Evolutionary Computation David B. Fogel, PhD Chief Scientist, Natural Selection, Inc., La Jolla Evolutionary computation encompasses a broad field of optimization algorithms that can be applied to diverse, difficult real-world problems. It is particularly useful in addressing stochastic, nonlinear, and time-varying optimization problems, including those arising in financial engineering. This tutorial will provide background on the inspiration, history, and the practical application of evolutionary computation to problems typical of those encountered in financial engineering. ---------------------------------------------------------------------------- Models for Stochastic Volatility: Some Recent Developments Nuno Cato Professor, New Jersey Institute of Technology, Newark Pedro J. F. de Lima Professor, The Johns Hopkins University, Baltimore In this tutorial, we will firstly discuss the importance of modeling stock market's volatility. Secondly, we will review the basic properties of GARCH- type and SV-type models and some of their most successful extensions, namely the SWitching ARCH (SWARCH) models. The performance of these models will be illustrated with some real data examples. Thirdly, we will discuss some problems with the estimation of these models and with their use for risk forecasting. Fourthly, we will describe some recent research and some novel extensions to these models, such as the Long-Memory Stochastic Volatility (LMSV) and the SWitching Stochastic Volatility (SWSV) models. By using examples from recent stock market behavior we illustrate the capabilities and shortcomings of these new modeling and forecasting tools. ---------------------------------------------------------------------------- From tlindroo at bennet.dc.tutech.fi Tue Feb 25 04:31:38 1997 From: tlindroo at bennet.dc.tutech.fi (Tommi Lindroos) Date: Tue, 25 Feb 1997 11:31:38 +0200 Subject: EANN '97: Invitation for participation Message-ID: <199702250931.LAA35528@bennet.dc.tutech.fi> Sorry for this unsolicited mail. It is being sent to you because you are apparently working in or interested in the field of neural networks. Most of these addresses are taken from neural network societies, and participants of earlier EANN conferences. Please let us know if your address should not be in here. International Conference on Engineering Applications of Neural Networks (EANN '97) Royal Institute of Technology Stockholm, Sweden 16-18 June 1997 INVITATION FOR PARTICIPATION AND PROGRAM OUTLINE The International Conference on Engineering Applications of Neural Networks (EANN '97) is the third conference in the series. The two earlier ones were held near Helsinki in 1995 and in London in 1996. The conference is a forum for presenting the latest results on neural network applications in technical fields. Over a hundred papers from 27 countries have been accepted for oral and poster presentations after a review of the abstracts. Some more information on the conference EANN '97 is available on the world wide web site at http://www.abo.fi/~abulsari/EANN97.html, and on last two conferences at http://www.abo.fi/~abulsari/EANN95.html and EANN96.html Contact details E-mail address : eann97 at kth.se Address : EANN '97 SEA PL 953, FIN 20101 Turku 10, Finland The conference has been organised with cooperation from Systems Engineering Association AB Nonlinear Solutions OY and Royal Institute of Technology Conference chairmen: Hans Liljenstrm and Abhay Bulsari Conference venue: Department of Numerical Analysis (NADA), Royal Institute of Technology (KTH), Osquars backe 2, 10044 Stockholm Registration information The conference fee will be SEK 4148 (SEK 3400 excluding VAT) until 28 February, and SEK 4978 (SEK 4080 excluding VAT) after that. The conference fee includes attendance to the conference and the proceedings. If your organisation (university or company or institute) has a VAT registration from a European Union country other than Finland, then your VAT number should be mentioned on the bank transfer as well as the registration form, and VAT need not be added to the conference fee. The correct conference fee amount should be received in the account number 207 799 342, Svenska Handelsbanken International, Stockholm branch. It can be paid by bank transfer, with all expenses paid by the sender, to "EANN Conference". To avoid extra bureaucracy and correction of the amount at the registration desk, make sure that you have taken care of the bank transfer fees. It is essential to mention the name of the participant with the bank transfer. If you need to pay it in another way (bank drafts, Eurocheques, postal order; no credit cards), please contact us at eann97 at kth.se. Invoicing will cost SEK 100. The tentative program outline is as on the following page. The detailed program will be prepared in the end of April. PROGRAM OUTLINE Sunday, 15 June 1600-1800 Registration Room E1 Room E2 Monday, 16 June 0830 Opening 0845 Vision (1) Control Systems (1) 1200 --- lunch break --- 1330 Vision (2) Control Systems (2) 1630 Discussion session Discussion session on Control on Vision Tuesday, 17 June 0830 Process Engineering Biomedical Engineering 1130 --- lunch break --- 1300 Metallurgy Mechanical Engineering 1530 Industrial Panel Discussion Wednesday, 18 June 0830 Hybrid systems Special applications 1130 --- lunch break --- 1300 Electrical/Electronics General Applications 1800 Closing In addition, there will be a poster session on Wednesday, and possibly an evening program on Tuesday. The indicated times are approximate and changes are still possible. Coffee breaks are not indicated here. International Conference on Engineering Applications of Neural Networks (EANN '97) Stockholm, Sweden 16-18 June 1997 Registration form Surname First Name Affiliation (name of the university/company/organisation) E-mail address Postal Address City Country Fax Have you submitted one or more abstracts ? Y/N Abstract number(s) Registration fee sent (amount) SEK ____________ by bank transfer number _______________ from Bank ______________ VAT registration number Date registration fee sent Date registration form sent Any special requirements ? --- END --- From karaali at ukraine.corp.mot.com Tue Feb 25 15:24:44 1997 From: karaali at ukraine.corp.mot.com (Orhan Karaali) Date: Tue, 25 Feb 1997 14:24:44 -0600 Subject: Neural Network Intern For Speech Recognition Message-ID: <199702252024.OAA18339@fiji.mot.com> NEURAL NETWORK INTERN FOR SPEECH RECOGNITION Motorola's Chicago Corporate Research Laboratories is currently seeking a Ph.D. student to join the Speech Synthesis and Machine Learning Group as an intern in its Speech Processing Systems Research Laboratory in Schaumburg, Illinois, for spring 1997. The internship will last at least three months. The Speech Synthesis and Machine Learning Group has developed innovative neural network and signal processing technologies for speech synthesis and speech recognition applications. The intern will work on a component of an HMM/neural network hybrid speech recognizer. The duties of the position include applied research, software development, and conducting experiments with speech data sets. Innovation in research, application of technology and a high level of motivation is the standard for all members of the team. The individual should be in an advanced stage of a Ph.D. program in EE, CS or a related discipline. The ability to work within a group to quickly implement and evaluate algorithms in a rapid research/development cycle is essential. Strong programming skills in the "C" language and solid knowledge of neural networks are required. Background in any of speech processing, statistical techniques, decision trees, and genetic algorithms is highly desirable. Please send text-format resume and cover letter to Orhan Karaali, karaali at mot.com. Motorola is an equal opportunity/affirmative action employer. We welcome and encourage diversity in our workforce. From ejua71 at tattoo.ed.ac.uk Wed Feb 26 12:57:35 1997 From: ejua71 at tattoo.ed.ac.uk (J A Bullinaria) Date: Wed, 26 Feb 97 17:57:35 GMT Subject: Conference announcement: NCPW4 Message-ID: <9702261757.ab24625@uk.ac.ed.tattoo> 4th Neural Computation and Psychology Workshop Connectionist Representations : Theory and Practice University of London, England Wednesday 9th April - Friday 11th April 1997 We have recently solicited, recieved, reviewed and accepted 36 abstracts for presentation as talks at this workshop. We now invite additional participants to register, attend and listen to these talks. AIMS AND OBJECTIVES This workshop is the fourth in a series, following on from the first at the University of Wales, Bangor (with theme "Neurodynamics and Psychology"), the second at the University of Edinburgh, Scotland ("Memory and Language") and the third at the University of Stirling, Scotland ("Perception"). The general aim is to bring together researchers from such diverse disciplines as artificial intelligence, applied mathematics, cognitive science, computer science, neurobiology, philosophy and psychology to discuss their work on the connectionist modelling of psychology. This years workshop is being hosted jointly by members of the Psychology Departments of Birkbeck College London and University College London. As in previous years there will be a theme to the workshop. We think that this years theme is sufficiently wide ranging and important that researchers in all areas of Neural Computation and Psychology will find it relevant and have something to say on the subject. The theme is to be: "Connectionist Representations : Theory and Practice". This covers many important issues ranging from the philosophical (such as the grounding problem) to the physiological (what can connectionist representations tell us about real neural systems) to the technical (such as what is necessary to get specific models to work). As in previous years we aim to keep the workshop fairly small, informal and single track. As always, participants bringing expertise from outside the UK are particularly welcome. SPEAKERS INCLUDE: Roland Baddeley (Oxford) Dennis Norris (APU Cambridge) Tony Browne (Mid Kent) Mike Page (APU Cambridge) Neil Burgess (UCL) Malti Patel (Sydney, Australia) Morten Christiansen (S. Calif.) Tim Shallice (UCL) Robert French (Liege, Belgium) Leslie Smith (Stirling) Peter Hancock (Stirling) John G. Taylor (King's London) Glyn Humphreys (Birmingham) Chris Thornton (Sussex) Geoff Goodhill (Georgetown, DC) Janet Vousden (Warwick) Our web site has a complete listing of all the talks and abstracts. REGISTRATION, FOOD AND ACCOMMODATION The workshop will be held in University College London, which is situated in the centre of London, near the British Museum and within easy walking distance of the West End and many of London's major attractions. The conference registration fee (which includes lunch and morning and afternoon tea/coffee each day) is 60 pounds. A special conference dinner (optional) is planned for the Thursday evening costing 20 pounds. Accommodation can be arranged in student residences or in local hotels, according to budget. The conference/accommodation area is easily accessible by the London Underground system ("The Tube"), with direct lines from London Heathrow Airport and all the major intercity train stations. Full registration and accommodation information is available at the conference web site: "http://prospero.psychol.ucl.ac.uk/ncpw4/". ORGANISING COMMITTEE John Bullinaria (Birkbeck College London) Dave Glasspool (University College London) George Houghton (University College London) CONTACT DETAILS Workshop email address for all correspondence: ncpw4 at psychol.ucl.ac.uk Workshop web page: http://prospero.psychol.ucl.ac.uk/ncpw4/ John Bullinaria, NCPW4, Centre for Speech and Language, Department of Psychology, Birkbeck College, Malet Street, London WC1E 7HX, UK. Phone: +44 171 631 6330, Fax: +44 171 631 6587 Email: j.bullinaria at psyc.bbk.ac.uk Dave Glasspool, NCPW4, Department of Psychology, University College London, Gower Street, London WC1E 6BT, UK. Phone: +44 171 380 7777 Xtn. 5418. Fax: +44 171 436 4276 Email: d.glasspool at psychol.ucl.ac.uk George Houghton, NCPW4, Department of Psychology, University College London, Gower Street, London WC1E 6BT, UK. Phone: +44 171 380 7777 Xtn. 5394. Fax: +44 171 436 4276 Email: g.houghton at psychol.ucl.ac.uk From mel at quake.usc.edu Thu Feb 27 01:20:18 1997 From: mel at quake.usc.edu (Bartlett Mel) Date: Thu, 27 Feb 1997 14:20:18 +0800 Subject: UCSD/USC Joint Symposium on Neural Computation Message-ID: <9702272220.AA00232@quake.usc.edu> CALL FOR PRESENTATIONS --- 4th Annual Joint Symposium on Neural Computation --- Co-sponsored by Institute for Neural Computation University of California, San Diego and Biomedical Engineering Department and Neuroscience Program University of Southern California to be hosted at The University of Southern California University Park Campus Saturday, May 17, 1997 9:00 a.m. to 6:00 p.m. In 1994, the Institute for Neural Computation at UCSD hosted the first Joint Symposium on Neural Computation with Caltech that brought together students and faculty for a day of short presentations. This year USC will be the site for the fourth Symposium and will feature as keynote speaker: Prof. Irving Biederman Departments of Psychology and Computer Science and the Neuroscience Program University of Southern California "Shape Representation in Mind and Brain" Submissions will be open to members of the Computational Neuroscience community of Southern California. Given the larger constituency than in previous years, authors are invited to contribute 300 word abstracts, which will be reviewed by a program committee consisting of the USC organizers and representatives of the INC. The contributed program will consist of 15 minute oral presentations, and posters. Abstracts selected for presentation will be included in the final program. DEADLINE FOR RECEIPT OF ABSTRACTS: March 28, 1997 Submissions should be e-mailed or mailed to Linda Yokote using the form below. Notification of acceptance as oral or poster presentation will be e-mailed to authors by April 18, 1997. A proceedings of short papers will be published by the INC. Contributions to the proceedings, based on both oral and poster presentations, must be submitted no later than May 30, 1997 for timely publication to: Institute for Neural Computation, University of California San Diego, 9500 Gilman Drive DEPT 0523, La Jolla, California 92093-0523. As in previous years, authors will retain copyright to their papers, so that they may be resubmitted elsewhere. Registration and attendance at the Symposium is open to the public. USC Organizing Committee: Dr. Bartlett Mel - Biomedical Engineering Department mel at quake.usc.edu, http://quake.usc.edu/lnc.html Dr. Michael Arbib - Professor of Computer Science and Neurobiology Director of the USC Brain Project arbib at pollux.usc.edu, http:/www-hbp.usc.edu/HBP ---------------------------------------------------------------------------- ---------------------------------------------------------------------------- 1997 JSNC Submission Form - Return to: Linda Yokote yokote at bmsrs.usc.edu (e-mail submissions preferred) US mail: Joint Symposium Biomedical Engineering Department USC, MC 1451 Los Angeles, CA 90089 (213)740-0840, (213)740-0343 fax ---------------------------------------------------------------------------- _____ I would like to attend the Symposium. Registration fee of $25 includes lunch and Proceedings. Checks are payable to the Department of Biomedical Engineering, USC. _____ I would like to give a presentation Title: _______________________________________________________________________ _______________________________________________________________________ Abstract: (300 word abstract goes here) _______________________________________________________________________ Speaker's Name: ______________________________________________________________ Affiliation/Department: ______________________________________________________ Address: _____________________________________________________________________ _____________________________________________________________________ _____________________________________________________________________ Telephone: ________________________ E-mail Address: _____________________ Others who should be listed as co-authors: ______________________________ ______________________________________________________________________________ ______________________________________________________________________________ ______________________________________________________________________________ --- DEADLINE FOR RECEIPT OF ABSTRACTS: March 28, 1997 --- From bishopc at helios.aston.ac.uk Thu Feb 27 05:09:52 1997 From: bishopc at helios.aston.ac.uk (Prof. Chris Bishop) Date: Thu, 27 Feb 1997 10:09:52 +0000 Subject: Isaac Newton Institute Message-ID: <3914.199702271009@sun.aston.ac.uk> Isaac Newton Institute NEURAL NETWORKS AND MACHINE LEARNING A six month programme at the Isaac Newton Institute for Mathematical Sciences, Cambridge, U.K. July to December 1997 Organisers: C M Bishop (Aston), D Haussler (UCSC), G E Hinton (Toronto), M Niranjan (Cambridge), L G Valiant (Harvard) The Isaac Newton Institute for Mathematical Sciences is an international research centre, sponsoring visitor programmes in topics across the whole spectrum of mathematical research. This programme will be an important international event in the field of neural computing. At any one time there will be around 20 to 25 long-term participants, as well as larger numbers of short-term visitors. Seminars are open to all, although long-term participation will be by invitation. Three major conferences are planned to take place during the programme: 1) "Generalization in Neural Networks and Machine Learning" (4 to 15 August). This will be a NATO Advanced Study Institute. 2) "Probabilistic Graphical Models" (1 to 5 September). 3) "Bayesian Methods" (15 to 19 December). In addition, there will be a number of workshops, generally of one week duration. Provisional themes for these include the following: "Learning in computer vision" (6 to 10 October), "HMM hybrid models and protein/DNA modelling" (20 to 24 October), "Applications" (3 to 7 November), "Non-stationarity" (10 to 14 November), "Information geometry" (8 to 12 December). Information about the programme, and a list of participants, can be found at the programme web site: http://www.newton.cam.ac.uk/programs/nnm.html General information about the Newton Institute can be found at: http://www.newton.cam.ac.uk/ To be kept informed of future developments, you can subscribe to the programme information mailing list by sending an e-mail to majordomo at newton.cam.ac.uk with a message whose body contains the line subscribe nnm-list For programme-specific enquires please contact: Christopher M. Bishop Neural Computing Research Group Dept. of Computer Science and Applied Mathematics Aston University, Birmingham B4 7ET, U.K. Tel. +44/0 121 333 4631 Fax. +44/0 121 333 4586 C.M.Bishop at aston.ac.uk From movellan at ergo.ucsd.edu Thu Feb 27 19:31:44 1997 From: movellan at ergo.ucsd.edu (Javier R. Movellan) Date: Thu, 27 Feb 1997 16:31:44 -0800 Subject: TR Announcement Message-ID: <199702280031.QAA29932@ergo.ucsd.edu> The following technical report is available online at http://cogsci.ucsd.edu (follow links to Tech Reports & Software ) Physical copies are also available (see the site for information). From Contour Completion to Image Schemas: A Modern Perspective on Gestalt Psychology. Adrian Robert (Communicated by Martin I. Sereno) Department of Cognitive Science University of California San Diego The Gestalt approach to psychology represents an early but comprehensive and systematic attempt to relate psychological and neural functioning. When the approach was first formulated and actively researched, however, too little was known about brain function to forge a precise and direct connection. As a result, the approach never fulfilled its initial promise of a rigorously founded psychology grounded in physical science and has fallen out of the favor and attention of most contemporary students of the mind. In this paper we re-examine Gestalt psychology with reference to what is currently known of dynamic mechanisms of brain function, particularly by exploring plausible neural substrates of perceptual grouping. We suggest, based on this examination, that although many of the details of the Gestalt proposals are in need of revision, the approach remains fundamentally viable, and the elegant character of its grounding and systematicity make it a valuable framework for organizing present knowledge at both neural and functional levels. From gerda at ai.univie.ac.at Fri Feb 28 04:00:58 1997 From: gerda at ai.univie.ac.at (Gerda Helscher) Date: Fri, 28 Feb 1997 10:00:58 +0100 Subject: New Book: NN and a new AI Message-ID: <33169ECA.2781E494@ai.univie.ac.at> !!! New Book Announcement !!! ================================================= Neural Networks and a New Artificial Intelligence ================================================= edited by Georg Dorffner International Thomson Computer Press, London, 1997 ISBN 1-85032-172-8 About the book: =============== Since the re-birth of interest in artificial neural networks in the mid 1980s, they have become a much-discussed topic, particularly in terms of their real contribution to the explanation and modelling of cognition, as part of the field of artificial intelligence (AI). This edited collection brings together a selection of papers from experts in their field, outlining the concrete contribution that neural computing has made to AI. "Neural Networks and a New Artificial Intelligence" is a collection of arguments, examples and critical elaborations from different views on how and whether neural networks can not only contribute to a better artificial intelligence, but can also revolutionise it by forming the basis for a truly alternative paradigm. Contents: ========= Introduction Part I: General topics - new AI as a whole New AI: naturalness revealed in the study of artificial intelligence (by Erich Prem, Vienna) Representational eclecticism - a foundation stone for the new AI? (by Chris Thornton, Brighton) Part II: concrete approaches and research strategies Towards a connectionist model of action sequences, active vision and breakdowns (by Hugues Bersini, Brussels) Complete autonomous systems: a research strategy for cognitive science (by Rolf Pfeifer and Paul Verschure, Zurich) Radical connectionism - a neural bottom-up approach to AI (by Georg Dorffner, Vienna) Connectionist explanation: taking position in the mind-brain dilemma (by Paul Verschure, Zurich) On growing intelligence (by Jari Vaario, Kyoto, and Setsuo Ohsuga, Tokyo) Part III: Issues in modelling high-level cognition Systematicity and generalization in compositional connectionist representations (by Lars Niklasson, Sk"ovde, and Noel Sharkey, Sheffield) Constructive learning in connectionist semantic networks (by Joachim Diederich and James Hogan, Brisbane) A connectionist Model for the interpretation of metaphors (by Stefan Wermter, Hamburg, and Ruth Hannuschka, Dortmund) Some issues in neural cognitive modelling (by Max Garzon, Memphis) Neural networks and a new AI - questions and answers (with contributions from all the authors) Index From jan at uran.informatik.uni-bonn.de Mon Feb 3 11:21:44 1997 From: jan at uran.informatik.uni-bonn.de (Jan Puzicha) Date: Mon, 3 Feb 1997 17:21:44 +0100 (MET) Subject: Preprints and Abstracts available online Message-ID: <199702031621.RAA02383@thalia.informatik.uni-bonn.de> This message has been posted to several lists. Sorry, if you receive multiple copies. The following six PREPRINTS are now available as abstracts and compressed postscript online via the WWW-Home-Page http://www-dbv.cs.uni-bonn.de/ of the |---------------------------------------------| |Computer Vision and Pattern Recognition Group| | of the University of Bonn, | | Prof. J. Buhmann, Germany. | |---------------------------------------------| 1.) Thomas Hofmann and Joachim Buhmann, Pairwise Data Clustering by Deterministic Annealing, IEEE Transactions on Pattern Analysis and Machine Intelligence PAMI 19(1), 1997. 2.) Joachim Buhmann and Thomas Hofmann, Robust Vector Quantization by Competitive Learning. In: Proceedings of the International Conference on Accoustics, Speech and Signal Processing (ICASSP'97), Munich, 1997. 3.) Hansjrg Klock and Joachim Buhmann, Multidimensional Scaling by Deterministic Annealing. In: Proceedings of the International Workshop on Energy Minimization Methods in Computer Vision and Pattern Recognition EMMCVPR'97, Venice. 4.) Thorsten Frhlinghaus and Joachim Buhmann. Real-Time Phase-Based Stereo for a Mobile Robot. in: Proceedings of the First Euromicro Workshop on Advanced Mobile robots. pp. 178-185, 1996. 5.) Thomas Hofmann, Jan Puzicha and Joachim M. Buhmann, Deterministic Annealing for Unsupervised Texture Segmentation. In: Proceedings of the International Workshop on Energy Minimization Methods in Computer Vision and Pattern Recognition (EMMCVPR'97), Venice. 6.) Thomas Hofmann, Jan Puzicha and Joachim M. Buhmann, A Deterministic Annealing Framework for Unsupervised Texture Segmentation. Technical Report IAI-TR-96-2, Institut fr Informatik III, University of Bonn. 1996. |--------------------------------------| | NEW: Test our algorithms on YOUR data| |--------------------------------------| If you want to test our unsupervised segmentation algorithms (see preprints 5. and 6. for more information) on your data, please download your data in any of the usual formats (pgm,gif,postscript,...) on our ftp-server uran.cs.uni-bonn.de in the directory /pub/dbv/incoming and write a short notice to jan at cs.uni-bonn.de. We will segment your data as soon as possible and send you the results. Feel free to have a look at the new online-presentation of texture segmentation results and the new links to related sites, conferences and jounals. If you have any questions or remarks, please let me know. Greetings Jan Puzicha -------------------------------------------------------------------- Jan Puzicha | email: jan at uran.cs.uni-bonn.de Institute f. Informatics III | jan at cs.uni-bonn.de University of Bonn | WWW : http://www.cs.uni-bonn.de/~jan | Roemerstrasse 164 | Tel. : +49 228 550-383 D-53117 Bonn | Fax : +49 228 550-382 -------------------------------------------------------------------- From rolf at rug104.cs.rug.nl Tue Feb 4 05:19:12 1997 From: rolf at rug104.cs.rug.nl (Rolf P. Wuertz) Date: Tue, 4 Feb 1997 11:19:12 +0100 (MET) Subject: New articles available from my website Message-ID: <199702041019.LAA02959@rug104.cs.rug.nl> The following articles are available online: --------------------------------------------------------------------- Rolf P. W"urtz. Neuronal theories and technical systems for face recognition. In Proceedings of the Fifth European Symposium On Artificial Neural Networks, Bruges (Belgium), 16-18 April 1997 ABSTRACT: I present various systems for the recognition of human faces. They consist of three steps: feature extraction, solving the correspondence problem, and the actual comparison with stored faces. Two of them are implemented in the Dynamic Link Architecture and are, therefore, close to biological hardware, the others are more technical in nature but also have some biological plausibility. At the end, I will briefly discuss the coherence with the results of psychophysical experiments on human face recognition. compressed PS source (350kB, 6 pages): http://www.cs.rug.nl/~rolf/esann97.ps.gz --------------------------------------------------------------------- Stephen McKenna, Sean Gong, Rolf P. W"urtz, Jonathan Tanner and Daniel Banin. Tracking facial feature points with Gabor wavelets and shape models. In Proceedings of the First International Conference on Audio- and Video-based Biometric Person Authentication Crans-Montana, Switzerland, 12-14 March 1997 ABSTRACT: A feature-based approach to tracking rigid and non-rigid facial motion is described. Feature points are characterised using Gabor Wavelets and can be individually tracked by phase-based displacement estimation. In order to achieve robust tracking a flexible shape model is used to impose global constraints upon the local feature points and to constrain the tracker. While there are many applications in facial analysis, the approach is quite generic and can be used for tracking other textured objects. compressed PS source (600 kB, 8 pages) http://www.cs.rug.nl/~rolf/gwt-pdm.ps.gz --------------------------------------------------------------------- Rolf P. W"urtz, Wolfgang Konen, and Kay-Ole Behrmann. How fast can neuronal algorithms match patterns? In Christoph von der Malsburg and Jan C. Vorbr"uggen and Werner von Seelen and Bernhard Sendhoff (eds.), Artificial Neural Networks - ICANN 96 , pages 145-150. Springer Verlag, 1996. ABSTRACT: We investigate the convergence speed of the Self Organizing Map (SOM) and Dynamic Link Matching (DLM) on a benchmark problem for the solution of which both algorithms are good candidates. We show that the SOM needs a large number of simple update steps and DLM a small number of complicated ones. A comparison of the actual number of floating point operations hints at an exponential vs. polynomial scaling behavior with increased pattern size. DLM turned out to be much less sensitive to parameter changes than the SOM. compressed PS source (60 kB, 6 pages) http://www.cs.rug.nl/~rolf/icann96.ps.gz +----------------------------------------------------------------------------+ | Rolf P. W"urtz | mailto: rolf at cs.rug.nl | URL: http://www.cs.rug.nl/~rolf/ | | Department of Computing Science, University of Groningen, The Netherlands | +----------------------------------------------------------------------------+ From wray at ptolemy-ethernet.arc.nasa.gov Wed Feb 5 01:12:06 1997 From: wray at ptolemy-ethernet.arc.nasa.gov (Wray Buntine) Date: Tue, 4 Feb 97 22:12:06 PST Subject: Research Scientist for autonomous data analysis Message-ID: <9702050612.AA01509@ptolemy.arc.nasa.gov> NASA's Center of Excellence in Information Technology at Ames Research Center invites candidates to apply for a position as Research Scientist in Information Technology: Position description: * We seek applicants to join a small team of space scientists and computer scientists in developing NASA's next generation smart spacecraft with on-board, autonomous data analysis systems. The group includes leading space scientists (Ted Roush, Virginia Gulick) and leading data analysts (Wray Buntine, Peter Cheeseman), and their counterparts at JPL. * The team is doing the research and development required for the task, and has a multi-year program with deliverables planned. This is not a pure research position, and requires dedication in seeing completion of the R&D milestones. * The applicant will be responsible for the information technology side of R&D, with guidance from senior space scientists on the project. * The research has strong links with on-going work at the Center of Excellence and is an integral part of NASA's long term goals. Candidate requirements: * Strong interest in demonstrating autonomous analysis systems to enhance science understanding in operational tests, with the ultimate goal of putting such systems in space. * Ph.D. degree in Computer Science, Electrical Engineering, or related field, and applied experience, possibly within the PhD. In exceptional cases, an M.S. degree with relevant work experience will suffice. * Knowledge of neural or probabilistic networks, machine learning, statistical pattern recognition, image processing, science data, processing, probabilistic algorithms, or related topics is essential. * Strong communication and organizational skills with the ability to lead a small team and interact with scientists. * Strong C programming and Unix skills (experimental, not necessarily production), with experience in programming mathematical algorithms: C++, Java, MatLab, IDL. Application deadline: * March 15th, 1997 (hardcopy required -- see below). Please send any questions by e-mail to the addresses below, and type "PI for Autonomous data analysis" as your header line. Dr. Ted Roush: roush at cloud9.arc.nasa.gov Dr. Wray Buntine: buntine at cloud9.arc.nasa.gov Full applications (which must include a resume and the names and addresses of at least two people familiar with your work) should be sent by surface mail (no e-mail, ftp or html applications will be accepted) to: Dr. Steve Lesh Attn: PI for Autonomous data analysis Mail Stop 269-1 NASA Ames Research Center Moffett Field, CA, 94025-1000 From hadley at cs.sfu.ca Wed Feb 5 18:14:16 1997 From: hadley at cs.sfu.ca (Bob Hadley) Date: Wed, 5 Feb 1997 15:14:16 -0800 (PST) Subject: an abstract to post Message-ID: <199702052314.PAA14631@melodious.cs.sfu.ca> Hi, I'd be grateful if someone could post this abstract on the Connectionist List. Thanks, -------------------------------------------------------------------- Robert F. Hadley (Bob) Phone: 604-291-4488 Associate Professor email: hadley at cs.sfu.ca School of Computing Science Simon Fraser University Burnaby, B.C. V5A 1S6 Canada ------------------------------------------------------------ FTP-host: ftp.fas.sfu.ca FTP-filename: /pub/cs/hadley/act.pas.ps The following paper is available by FTP and on the WWW. (instructions below) ------------------------------------- Acquisition of the Active-Passive Distinction from Sparse Input and No Error Feedback by Robert F. Hadley and Vlad C. Cardei School of Computing Science Simon Fraser University Burnaby, B.C. V5A 1S6 Canada hadley at cs.sfu.ca vcardei at cs.sfu.ca Technical Report -- CSS-IS TR 97-01 ABSTRACT A connectionist-inspired, parallel processing network is presented which learns, on the basis of (relevantly) sparse input, to assign meaning interpretations to *novel* test sentences in both active and passive voice. Training and test sentences are generated from a simple recursive grammar, but once trained, the network successfully processes thousands of sentences containing deeply embedded clauses. All training is unsupervised with regard to error feedback -- only Hebbian and Kohonen forms of training are employed. In addition, the active-passive distinction is acquired without any supervised provision of cues or flags (in the output layer) that indicate whether the input sentence is in active or passive sentence. In more detail: (1) The model learns on the basis of a corpus of about 1000 sentences while the set of potential test sentences contains over 100 million sentences. (2) The model generalizes its capacity to interpret active and passive sentences to substantially deeper levels of clausal embedding. (3) After training, the model satisfies criteria for strong syntactic and strong semantic *systematicity* that humans also satisfy. (4) Symbolic message passing occurs within the model's output layer. This symbolic aspect reflects certain prior *language acquistion* assumptions. (60 pages, 1.2 spacing) ----------------------------------------------------------- You can obtain the above paper via ftp by doing the following: ftp ftp.fas.sfu.ca When asked for your name, type the word: anonymous When asked for a password, use your e-mail address. Then, you should change directory as follows: cd pub/cs/hadley and then do a get, as in: get act.pas.ps To exit from ftp, type : quit .................. Also, available using web brousers at: http://fas.sfu.ca/cs/people/GradStudents/vcardei/personal/Projects/ From rybaki at eplrx7.es.dupont.com Wed Feb 5 15:20:07 1997 From: rybaki at eplrx7.es.dupont.com (Ilya Rybak) Date: Wed, 5 Feb 1997 15:20:07 -0500 Subject: information about WEB cites, please post Message-ID: <199702052020.PAA15815@pavlov> Dear colleagues, The WEB page of our models of respiratory pattern generation has been updated and now includes abstracts of recent publications. One of main objectives of this work was to consider integration of intrinsic cellular (e.g. kinetics of ionic channels , Ca dynamics) and network propitious , and the role of this integration in generation and control of the respiratory pattern. The address of the page is: http://www.voicenet.com/~rybak/resp.html The WEB page of BMV model (Behavioral Model of Visual perception and invariant recognition) has been also significantly updated. The model is developed on the basis of some principles of biological vision and is able to recognize complex images (scene objects, faces, etc.) invariantly in respect to shift, rotation and scale. A DEMO and recent paper may be downloaded from the page. The address is: http://www.voicenet.com/~rybak/vnc.html Interested people are welcome to visit both cites. (Please be patient, it takes some time to get the page first time). Best wishes, Ilya Rybak DuPont Central Research rybaki at eplrx7.es.dupont.com http://www.voicenet.com/~rybak/ From isbell at ai.mit.edu Thu Feb 6 10:21:17 1997 From: isbell at ai.mit.edu (Charles L Isbell) Date: Thu, 6 Feb 1997 10:21:17 -0500 (EST) Subject: MIMIC NIPS*96 paper available for ftp Message-ID: <199702061521.KAA25200@panther.ai.mit.edu> Recently, Shumeet Baluja announced on this list the availability of CMU-CS-97-107, "Using Optimal Dependency-Trees for Combinatorial Optimization: Learning the Structure of the Search Space" by Baluja and Davies. That paper discusses and extends some work presented in "MIMIC: Finding Optima by Estimating Probability Densities" by De Bonet, Isbell, and Viola (to appear in NIPS*96). It seems worthwhile to mention that this paper is also now available: ftp://ftp.ai.mit.edu/pub/isbell/mimic.ps.gz Abstract In many optimization problems, the structure of solutions reflects complex relationships between the different input parameters. For example, experience may tell us that certain parameters are closely related and should not be explored independently. Similarly, experience may establish that a subset of parameters must take on particular values. Any search of the cost landscape should take advantage of these relationships. We present MIMIC, a framework in which we analyze the global structure of the optimization landscape. A novel and efficient algorithm for the estimation of this structure is derived. We use knowledge of this structure to guide a randomized search through the solution space and, in turn, to refine our estimate of the structure. Our technique obtains significant speed gains over other randomized optimization procedures. Peace. From A.Sharkey at dcs.shef.ac.uk Thu Feb 6 05:49:56 1997 From: A.Sharkey at dcs.shef.ac.uk (Amanda Sharkey) Date: Thu, 6 Feb 97 10:49:56 GMT Subject: IEE Colloquium, Neural Networks for Industrial Applications Message-ID: <9702061049.AA25425@gw.dcs.shef.ac.uk> Announcing, IEE COLLOQUIUM: NEURAL NETWORKS FOR INDUSTRIAL APPLICATIONS. Wednesday February 12, 1997 at the Institution of Electrical Engineers, Savoy Place, London, WC2R OBL. Chairperson: Amanda Sharkey (University of Sheffield) Speakers: L. Tarassenko (University of Oxford, UK), 'Logicook and QUESTAR: two case studies in successful technology transfer' G. Hesketh (AEA Technology, plc, UK), 'COUNTERMATCH: a neural network approach to automatic signal verification' F. Fogelman Soulie (Sligos, France), 'Prediction in banking applications' H.H. Thodberg (Thodberg Scientific Computing, Denmark), 'The automation of intuitive visual expertise: a classification of beef carcasses using neural networks.' V. Tresp (Siemens AG, Germany), 'Neural Networks at Siemens: products and perspectives.' P. Cowley (Rolls-Royce plc, UK), 'Condition Monitoring high integrity systems.' A.J. Morris (University of Newcastle upon Tyne, UK) 'Building robust neural networks for industrial process.' Colloquium organised by Professional Group C4 (Artificial Intelligence) and co-sponsored by the British Computer Society Specialist Group on Expert Systems, the Society for the Study of Artificial Intelligence and Simulation of Human Behaviour (AISB), and the Neural Computing Applications Forum (NCAF). For information about registration, contact UK 0171 240 1871 Ext: 2205/6 From mkearns at research.att.com Thu Feb 6 15:25:58 1997 From: mkearns at research.att.com (Michael J. Kearns) Date: Thu, 6 Feb 1997 15:25:58 -0500 (EST) Subject: Paper on leave-one-out cross validation available Message-ID: <199702062025.PAA09030@radish.research.att.com> The following paper is now available in compressed postscript format at http://www.research.att.com/~mkearns under the "Model Selection/Complexity Regularization" section of my publications. Algorithmic Stability and Sanity-Check Bounds for Leave-One-Out Cross-Validation Michael Kearns (AT&T Labs) and Dana Ron (MIT) In this paper we prove "sanity-check" bounds for the error of the leave-one-out cross-validation estimate of the generalization error: that is, bounds showing that the worst-case error of this estimate is not much worse than that of the training error estimate. The name sanity-check refers to the fact that although we often expect the leave-one-out estimate to perform considerably better than the training error estimate, we are here only seeking assurance that its performance will not be considerably worse. Perhaps surprisingly, such assurance has been given only for rather limited cases in the prior literature on cross-validation. Any nontrivial bound on the error of leave-one-out must rely on some notion of algorithmic stability. Previous bounds relied on the rather strong notion of hypothesis stability, whose application was primarily limited to nearest-neighbor and other local algorithms. Here we introduce the new and weaker notion of error stability, and apply it to obtain sanity-check bounds for leave-one-out for other classes of learning algorithms, including training error minimization procedures and Bayesian algorithms. We also provide lower bounds demonstrating the necessity of error stability for proving bounds on the error of the leave-one-out estimate, and the fact that for training error minimization algorithms, in the worst case such bounds must still depend on the Vapnik-Chervonenkis dimension of the hypothesis class. From joachim at fit.qut.edu.au Fri Feb 7 00:48:56 1997 From: joachim at fit.qut.edu.au (Joachim Diederich) Date: Fri, 7 Feb 1997 15:48:56 +1000 (EST) Subject: PhD Scholarships Message-ID: <199702070548.PAA06756@aldebaran.fit.qut.edu.au> NEUROCOMPUTING RESEARCH CENTRE QUEENSLAND UNIVERSITY OF TECHNOLOGY (QUT) BRISBANE, AUSTRALIA FOUR PhD SCHOLARSHIPS QUT is offering PhD scholarships in the area of Neurocomputing and Artificial Intelligence. Research activities will include natural language acquisition, grammar induction and connectionist forms of analytical learning. The scholarships include an annual tax-free living allowance as well as the tuition fees for students from outside Australia/New Zealand. The successful applicants are expected to have a four-year degree in Computer Science/Cognitive Science (e.g. upper- levels Honours or Masters). Good knowledge of artificial neural networks and symbolic machine learning is essential; good programming skills are desirable. The scholars will work in the Neurocomputing Research Centre and will be part of a high-profile team comprising senior researchers and other PhD students. The centre offers state-of-the-art computing facilities including networked Unix workstations plus support for national and international conference travel. Potential applicants should contact Professor Joachim Diederich, Neurocomputing Research Centre, e-mail: joachim at fit.qut.edu.au. Enquiries should include a brief resume plus a reference to the scholarship of interest. These positions will be filled immediately. SCHOLARSHIP 1 NEUROCOMPUTING AND NATURAL LANGUAGE LEARNING QUT is offering a PhD scholarship in neurocomputing and natural language learning as part of a collaborative project with the University of California, Berkeley and the Australian National University. Research activities will include theoretical and experimental aspects of Connectionist Natural Language Processing, in particular, language learning. A strong background in computer science is required and qualifications in linguistics are an advantage. SCHOLARSHIP 2 COMPUTATIONAL INTELLIGENCE FOR THE PRIMARY INDUSTRIES This project investigates computerised techniques for the support of primary industies. In particular, statistical methods, neurocomputing and artificial intelligence techniques are evaluated. The comparison will be done by methods such as cross-validation and bootstrap, as well as by the use of "real world scenarios." The objective is an integrated information system which combines these techniques and aims at a performance which cannot be achieved by any of the above mentioned methods in isolation. SCHOLARSHIP 3 RECURRENT NEURAL NETWORKS AND CONTEXT-FREE LANGUAGES Various kinds of simple recurrent networks (SRNs) and their learning strategies will be evaluated and compared towards the implementation of a recogniser of context free languages modelled on a pushdown automaton. Starting with a Giles- style higher order network at the core, it is expected to devolve into a higher order recurrent network and a number of SRNs each with clearly delineated responsibilities. SCHOLARSHIP 4 CONNECTIONIST SYSTEMS AND ANALYTICAL LEARNING This project will focus on connectionists forms of analytical learning such as explanation-based generalisation. Connectionist systems for the representation of structured knowledge will be used for deduction and learning. Neurocomputing Research Centre Queensland University of Technology Box 2434, Brisbane Q 4001 AUSTRALIA Phone: +61 7 3864-2143 Fax: +61 7 3864-1801 http://www.fit.qut.edu.au/NRC/ From ucganlb at ucl.ac.uk Thu Feb 6 12:18:30 1997 From: ucganlb at ucl.ac.uk (Dr Neil Burgess - Anatomy UCL London) Date: Thu, 06 Feb 97 17:18:30 +0000 Subject: Discussion Meeting on Spatial Cognition, March 97 Message-ID: <33070.9702061718@link-1.ts.bcc.ac.uk> ``What are the parietal and hippocampal roles in spatial cognition?'' Date: 19th and 20th March 1997 Location: The Royal Society, 6 Carlton House Terrace, London SW1 5AG, U.K. Organisers: Neil Burgess & John O'Keefe Synopsis: The parietal cortex has long been identified as the neural substrate for spatial behaviour. More recently a spatial role for the hippocampus has been postulated, particularly in rats. Do these areas both perform the same computations, do they complement each other, are they both nodes in an overall network of spatial processing, or do they just have different functions in different species? These questions will be addressed with respect to neurophysiology, neuropsychology and computational modelling. Other Info: Royal Society discussion meetings are free to attend, but an attendance form should be returned to the Royal Society by the 10th March (lunch tickets cost 11.00 pounds sterling). Enquiries and requests for forms should be made to: Science Promotion Section, The Royal Society, 6 Carlton House Terrace, London SW1 5AG, U.K. Tel: +44 (0)171 839 5561 x2574, Fax: +44 (0)171 451 2693 Speakers: AMARAL, David (University of California) ANDERSEN, Richard (California Institute of Technology) ARBIB, Michael (University of Southern California) BERTHOZ, Alain (CNRS, Paris) BURGESS, Neil (University College London) GAFFAN, David (University of Oxford) KARNATH, Hans-Otto (University of Tubingen) MAGUIRE, Eleanor (Institute of Neurology, London) McNAUGHTON, Bruce (University of Arizona) MILNER, Brenda (Montreal Neurological Hospital) MISHKIN, Mortimer (NIMH, Bethesda) MORRIS, Richard (University of Edinburgh) MULLER, Robert (SUNY Health Sciences Center, Brooklyn) O'KEEFE, John (University College London) VALLAR, Guiseppe (University of Rome) Programme: Wednesday 19th March Session I. Chair: John O'Keefe (University College London) 9.45 O'KEEFE, John (University College London) Introduction 10.00 AMARAL, David (University of California) A neuroanatomical analysis of sensory inputs via temporal and parietal regions to the rat and monkey hippocampal formation 10.40 coffee 11.10 VALLAR, Guiseppe (University of Rome) Spatial frames of reference and somatosensory processing: a neuropsychological perspective 11.50 KARNATH, Hans-Otto (University of Tubingen) Spatial orientation and the representation of space with parietal lobe lesions 12.30 lunch Session II. Chair: Brenda Milner (Montreal Neurological Hospital) 2.00 ANDERSEN, Richard (California Institute of Technology) Multimodal integration for stable representations of space in the posterior parietal cortex 2.40 ARBIB, Michael (University of Southern California) Modeling the Specialization of Parietal Subregions in Providing Visual Affordances for Diverse Tasks 3.20 tea 3.50 BERTHOZ, Alain (CNRS, Paris) Neural basis of spatial memory during real and imagined locomotor trajectories 4.30 MISHKIN, Mortimer (NIMH, Bethesda) Cortical sensory processing streams and the hippocampus 5.10 general discussion 5.20 close Thursday 20th March Session III. Chair: Mortimer Mishkin (NIMH, Bethesda) 10.00 MILNER, Brenda (Montreal Neurological Hospital) Medial temporal lobe contributions to object location memory 10.40 coffee 11.10 MAGUIRE, Eleanor (Institute of Neurology, London) Hippocampal involvement in human topographical memory 11.50 GAFFAN, David (University of Oxford) Episodic memory, neglect and hemiamnesia 12.30 lunch Session IV. Chair: Lynn Nadel (University of Arizona) 2.00 MORRIS, Richard (University of Edinburgh) Hippocampal plasticity : synaptic changes underlying the automatic recording of attended experience 2.40 MULLER, Robert (SUNY Health Sciences Center, Brooklyn) A Topological approach to cognitive mapping 3.20 tea 3.50 McNAUGHTON, Bruce (University of Arizona) Memory reprocessing in cortico-cortico and hippocampo-cortical neuronal ensembles 4.30 BURGESS, Neil (University College London) Robotic and neuronal simulation of the hippocampus and rat navigation 5.10 general discussion 5.20 round-up 5.30 Close From wyler at iam.unibe.ch Fri Feb 7 11:01:39 1997 From: wyler at iam.unibe.ch (Kuno Wyler) Date: Fri, 7 Feb 1997 17:01:39 +0100 Subject: POSTDOCTORAL RESEARCH FELLOWSHIP available Message-ID: <9702071601.AA06468@garfield.unibe.ch> POSTDOCTORAL RESEARCH FELLOWSHIP -------------------------------- Neural Computing Research Group Institute of Informatics and Applied Mathematics University of Bern, Switzerland The Neural Computing Research Group at the University of Bern is looking for a highly motivated individual for a two year postdoctoral research position as project manager in the area of development of a neuromorphic perception system based on multi sensor fusion. The aim of the project is to develop a neurobiologically plausible perception system for novelty detection in a real world environment (e.g. quality control in industrial production lines or supervision of security zones) using the fusion of audio and video information and fast learning algorithms for the dynamic binding of signal features. Potential candidates should have good skills in applied optical sciences with a sound background in mathematics and signal processing, and working knowledge in neurobiology and neural networks. Further knowledge of programming (Matlab, LabView or C/C++) or VLSI technology is highly desirable but not required. Project management includes the supervision and the scientific leading of the collaborators (1 Postdoc/ 2 PhD candidates). The position will begin May 1, 1997, with possible renewal for an additional two years. The initial salary is SFr. 70'000/year (approx. $50'000). To apply for this position, send your curriculum vitae, publication list with one or two sample publications and two letters of reference before March 15, 1997, by surface mail to (applications by e-mail will not be accepted) Neural Computing Research Group Postdoctoral Research Fellowship attn. Mrs. M. Aberegg Institute of Informatics and Applied Mathematics University of Bern Neubrueckstrasse 10 CH-3012 Bern Switzerland For further information you can send e-mail to wyler at iam.unibe.ch or mueller at iam.unibe.ch. From reza at bme.jhu.edu Fri Feb 7 12:40:39 1997 From: reza at bme.jhu.edu (Reza Shadmehr) Date: Fri, 7 Feb 1997 12:40:39 -0500 (EST) Subject: faculty position in computational neuroscience Message-ID: <199702071740.MAA17245@reflex.bme.jhu.edu> JOHNS HOPKINS UNIVERSITY BIOMEDICAL ENGINEERING The Department of Biomedical Engineering at the Johns Hopkins University invites applications for a tenured or tenure-track faculty position in computational neuroscience. The successful candidate must be able to establish an independently funded research program that builds on or complements existing research programs. The candidate will also participate in the teaching of undergraduate and graduate students. For more information on our department see: http://www.bme.jhu.edu. Interested applicants should send their CV and names of three referees to: Dr. Murray B. Sachs Faculty Search Department of Biomedical Engineering 710 Traylor Building 720 Rutland Avenue Baltimore, MD 21205 The Johns Hopkins University is an Equal Opportunity/Affirmative Action Institution. From Dave_Touretzky at DST.BOLTZ.CS.CMU.EDU Sun Feb 9 21:44:50 1997 From: Dave_Touretzky at DST.BOLTZ.CS.CMU.EDU (Dave_Touretzky@DST.BOLTZ.CS.CMU.EDU) Date: Sun, 09 Feb 97 21:44:50 EST Subject: CNBC summer undergraduate research program Message-ID: <5114.855542690@DST.BOLTZ.CS.CMU.EDU> The Center for the Neural Basis of Cognition, a joint project of Carnegie Mellon University and the University of Pittsburgh, is seeking applications from top-quality undergraduates interested in pursuing summer research in cognitive or computational neuroscience. The CNBC summer training program is a ten week intensive program of lectures, laboratory tours, and guided research. State of the art facilities include computerized microscopy; laboratories for human and animal electrophysiological recording; behavioral assessment laboratories for rat, primate, and human experimentation; MRI and PET scanners for brain imaging; the Pittsburgh Supercomputing Center; and a regional medical center providing access to human clinical populations. The Summer Training Program is a National Science Foundation sponsored program; we expect to support ten students in each of the next five years. Applications are encouraged from students with interests in biology, neuroscience psychology, engineering, physics, mathematics, computer science, or robotics. To be eligible, students must not yet have completed their bachelor's degree at the time they participate. The application deadline is March 15, 1997. For more information about the program, and detailed application instructions, see our web site at: http://www.cnbc.cmu.edu/Training/summer/index.html In addition to its summer program, the Center for the Neural Basis of Cognition (CNBC) offers an interdisciplinary training program for Ph.D. and postdoctoral students in collaboration with various affiliated departments at Carnegie Mellon University and the University of Pittsburgh. This training program is the descendant of the Neural Processes in Cognition program started in 1990 under the National Science Foundation. We now have thirty six graduate students and thirty five faculty affiliated with the CNBC. The program focuses on understanding higher level brain function in terms of neurophysiological, cognitive, and brain imaging data complemented with computational modeling. Individually designed programs of study encompass cellular and systems neuroscience, computational neuroscience, cognitive modeling, and brain imaging. For a brochure describing the graduate training program and application materials, contact us at the following address: Center for the Neural Basis of Cognition 115 Mellon Institute 4400 Fifth Avenue Pittsburgh, PA 15213 Telephone: (412) 268-4000 Fax: (412) 268-5060 Email: cnbc-admissions at cnbc.cmu.edu This material is also available on our web site at http://www.cnbc.cmu.edu Faculty: The CNBC training faculty includes: German Barrionuevo (Pitt Neuroscience): LTP in hippocampal slice; Marlene Behrmann (CMU Psychology): spatial representations in parietal cortex; Pat Carpenter (CMU Psychology): mental imagery, language, and problem solving; Jonathan Cohen (CMU Psychology): schizophrenia; dopamine and attention; Carol Colby (Pitt Neuroscience): spatial reps. in primate parietal cortex; Bard Ermentrout (Pitt Mathematics): oscillations in neural systems; Julie Fiez (Pitt Psychology): fMRI studies of language; John Horn (Pitt Neurobiology): synaptic learning in autonomic ganglia; Allen Humphrey (Pitt Neurobiology): motion processing in primary visual cortex; Marcel Just (CMU Psychology): visual thinking, language comprehension; Eric Klann (Pitt Neuroscience): hippocampal LTP and LTD; Alan Koretsky (CMU Biological Sciences): new fMRI techniques for brain imaging; Tai Sing Lee (CMU Comp. Sci.): primate visual cortex; computer vision; David Lewis (Pitt Neuroscience): anatomy of frontal cortex; James McClelland (CMU Psychology): connectionist models of cognition; Carl Olson (CNBC): spatial representations in primate frontal cortex; David Plaut (CMU Psychology): connectionist models of reading; Michael Pogue-Geile (Pitt Psychology): development of schizophrenia; John Pollock (CMU Biological Sci.): neurodevelopment of the fly visual system; Walter Schneider (Pitt Psychology): fMRI studies of attention and skill acquisition; Charles Scudder (Pitt Neurobiology): motor learning in cerebellum; Susan Sesack (Pitt Neuroscience): anatomy of the dopaminergic system; Dan Simons (Pitt Neurobiology): sensory physiology of the cerebral cortex; William Skaggs (Pitt Neuroscience): representations in rodent hippocampus; and David Touretzky (CMU Comp. Sci.): hippocampus, rat navigation, animal learning. Walter Schneider David Touretzky Professor of Psychology Computer Science Department University of Pittsburgh Carnegie Mellon University From meunier at orphee.polytechnique.fr Mon Feb 10 08:02:44 1997 From: meunier at orphee.polytechnique.fr (Claude Meunier) Date: Mon, 10 Feb 1997 14:02:44 +0100 Subject: No subject Message-ID: Research associate Center for Theoretical Physics Ecole Polytechnique, France The Center for Theoretical Physics invites applications for a possible one year position of research associate (no teaching duties), starting Fall 1997. Six month stays may also be considered. Candidates are expected to carry out their research in the framework of our existing research programs in: - computational neurosciences - condensed matter - field theory and particle physics - plasma physics Applicants should submit a resume with publications list, statement of research interests, and two letters of references to: Dr. M.-N. Bussac Centre de Physique Th?orique Ecole Polytechnique 91128 Palaiseau cedex France Fax: 00 33 01 69 33 30 08 E-mail: bussac at cpth.polytechnique.fr Send requests of information on this temporary position and post doctoral positions in theoretical physics at Ecole Polytechnique to the same address. The closing date for the receipt of applications is March 15, 1997. From klaus at prosun.first.gmd.de Mon Feb 10 13:22:12 1997 From: klaus at prosun.first.gmd.de (Klaus-R. Mueller) Date: Mon, 10 Feb 1997 19:22:12 +0100 Subject: Call for Tricks ;-) Message-ID: <9702101822.AA06771@prosun.first.gmd.de> ********************************************* * Call for Papers * * Tricks of the Trade * ********************************************* Dear Colleagues, At the Nips*96 workshops we had a workshop called "Tricks of the Trade: How to Make Algorithms Really Work". As a follow-up to this workshop, we are collecting papers for a "Book of Tricks" (this is our working title), which will be tentatively published in the LNCS State-of-the-Art Surveys series. We would like to invite you to contribute. What is a Trick?: A technique, rule-of-thumb, or heuristic that * is easy to describe and understand * can make a real difference in practice * is not (yet) part of well documented technique * has broad application and may or may not (yet) have a theoretical explanation. Content and Format: In order to keep everything focussed, we suggest the following main topics of the book: 1. architectural tricks 2. sampling and data preprocessing 3. speeding learning procedures 4. improving generalization and optimization To give consistency across papers, we would like there to be a structural similarity in the contents which is based on the outline we proposed for the talks: - intro/motivation - trick - where has it been tried and how well does it work? - why does it work? - possible theoretical explanation (if any), - heuristic explanation (if any) & discussion For More Details: Please look at the call for papers on our web site: In the US: http://www.willamette.edu/~gorr/tricks/guidelines.html In Europe: http://www.first.gmd.de/persons/Mueller.Klaus-Robert/CALL.html If you have questions, please email to: Jenny Orr (gorr at willamette.edu) Klaus-Robert M"uller (klaus at first.gmd.de) Rich Caruana (caruana at cs.cmu.edu) From meunier at orphee.polytechnique.fr Tue Feb 11 04:02:01 1997 From: meunier at orphee.polytechnique.fr (Claude Meunier) Date: Tue, 11 Feb 1997 10:02:01 +0100 Subject: research associate position Message-ID: Research associate Center for Theoretical Physics Ecole Polytechnique, France The Center for Theoretical Physics invites applications for a possible one year position of research associate (no teaching duties), starting Fall 1997. Six month stays may also be considered. Candidates are expected to carry out their research in the framework of our existing research programs in: - computational neurosciences - condensed matter - field theory and particle physics - plasma physics Applicants should submit a resume with publications list, statement of research interests, and two letters of references to: Dr. M.-N. Bussac Centre de Physique Th=E9orique Ecole Polytechnique 91128 Palaiseau cedex France fax: 00 33 01 69 33 30 08 E-mail: bussac at cpth.polytechnique.fr Send requests of information on this temporary position and post doctoral positions in theoretical physics at Ecole Polytechnique to the same address. The closing date for the receipt of applications is March 15, 1997. From karaali at ukraine.corp.mot.com Tue Feb 11 12:34:16 1997 From: karaali at ukraine.corp.mot.com (Orhan Karaali) Date: Tue, 11 Feb 1997 11:34:16 -0600 Subject: Motorola NN Speech Synthesizer Article Message-ID: <199702111734.LAA03748@fiji.mot.com> FTP-host: archive.cis.ohio-state.edu FTP-filename: /pub/neuroprose/karaali.synthesis_wcnn96.ps.Z Motorola Neural Network Speech Synthesizer Article A new neural network based speech synthesizer has been developed here at Motorola Chicago Corporate Research Laboratories by the Speech Synthesis and Machine Learning Group. We believe that the quality of the synthesized speech it produces surpasses the current state of the art, particularly in naturalness. An invited paper describing this neural network speech synthesizer was presented in the Speech Session of the World Congress on Neural Networks 96 in San Diego. This paper is now available in the NEUROPROSE archive as karaali.synthesis_wcnn96.ps.Z. If you have a problem getting the paper from NEUROPROSE, I can email it to you. Orhan Karaali email: karaali at mot.com --------------------------------------------------------------------- Speech Synthesis with Neural Networks Orhan Karaali, Gerald Corrigan, and Ira Gerson Motorola, Inc., 1301 E. Algonquin Road, Schaumburg, IL 60196 karaali at mot.com, corrigan at mot.com, gerson at mot.com ABSTRACT Text-to-speech conversion has traditionally been performed either by concatenating short samples of speech or by using rule-based systems to convert a phonetic representation of speech into an acoustic representation, which is then converted into speech. This paper describes a system that uses a time-delay neural network (TDNN) to perform this phonetic-to-acoustic mapping, with another neural network to control the timing of the generated speech. The neural network system requires less memory than a concatenation system, and performed well in tests comparing it to commercial systems using other technologies. ----- End Included Message ----- From lawrence at research.nj.nec.com Wed Feb 12 13:05:43 1997 From: lawrence at research.nj.nec.com (Steve Lawrence) Date: Wed, 12 Feb 1997 13:05:43 -0500 Subject: Student Position: Learning for Agents and Multi-Agent Systems Message-ID: <19970212130543.19266@purgatory.nj.nec.com> The NEC Research Institute in Princeton, NJ has an immediate opening for a student research position in the area of learning for agents and multi-agent systems. Candidates must have experience in research and be able to effectively communicate research results. Ideal candidates will have knowledge of one or more machine learning techniques (e.g. neural networks, decision trees, rule based systems, and nearest neighbor techniques), and be proficient in the software implementation of algorithms. NEC Research provides an outstanding research environment with many recognized experts and excellent resources including several multiprocessor machines. Interested applicants should apply by email, mail or fax including their resumes and any specific interests in learning for agents and multi-agent systems to: Dr. C. Lee Giles NEC Research Institute 4 Independence Way Princeton NJ 08540 Phone: (609) 951 2642 Fax: (609) 951 2482 Email: giles at research.nj.nec.com -- Steve Lawrence <*> http://www.neci.nj.nec.com/homepages/lawrence From wermter at nats5.informatik.uni-hamburg.de Wed Feb 12 07:45:25 1997 From: wermter at nats5.informatik.uni-hamburg.de (Stefan Wermter) Date: Wed, 12 Feb 1997 13:45:25 +0100 Subject: JAIR article - connectionist natural language learning Message-ID: <199702121245.NAA12800@nats6> A non-text attachment was scrubbed... Name: not available Type: x-sun-attachment Size: 4021 bytes Desc: not available Url : https://mailman.srv.cs.cmu.edu/mailman/private/connectionists/attachments/00000000/c4d77933/attachment-0001.ksh From beiu at lanl.gov Tue Feb 11 22:50:18 1997 From: beiu at lanl.gov (Valeriu Beiu) Date: Tue, 11 Feb 1997 20:50:18 -0700 Subject: CFP: IV Brazilian Symposium on Neural Networks SBRN'97 Message-ID: <2.2.32.19970212035018.003026f0@nis-pop.lanl.gov> /\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\ I V t h B R A Z I L I A N S Y M P O S I U M O N N E U R A L N E T W O R K S SBRN'97 /\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\ Goiania, GO, Brazil, December 03 - 05, 1997 Sponsored by: Brazilian Computer Society (SBC) International Neural Networks Society (INNS) Organized by the School of Electrical Engineering (EEE) Federal University of Goias (UFG) Conference Chairmen: Weber Martins (EEE-UFG) Dibio Leandro Borges (EEE-UFG) ======================================================================= Latest information will be placed at the conference WWW-page http://www.eee.ufg.br/sbrn97/ ======================================================================= P u r p o s e ----------------------------------------------------------------------- The fourth Brazilian Symposium on Neural Networks will be held at the School of Electrical Engineering, Federal University of Goias (UFG), Goiania, GO (Brazil), on 3rd, 4th, and 5th December, 1997. The Brazilian Symposia on Neural Networks are organized by the interest group in Neural Networks of the Brazilian Computer Society. In order to promote research in Neural Networks and Evolutionary Computation, and scientific interchange among Brazilian AI researchers and practitioners, and their counterparts worldwide, papers from the international community are most welcome. The papers will be reviewed by an international program committee. T o p i c s o f I n t e r e s t ----------------------------------------------------------------------- Submissions are invited on substantial, original, and previously unpublished research in all aspects of Neural Networks, including, but not limited to: * Biological Perspectives * Cognitive Modeling * Dynamic Systems * Evolutionary Computation * Fuzzy Logic * Hardware Implementation * Hybrid Systems * Learning Models * Neural Network Algorithms and Architectures * Neural Network Applications * Optimization * Pattern Recognition and Image Processing * Robotics and Control * Signal Processing * Theoretical Models P r o g r a m C o m m i t t e e ----------------------------------------------------------------------- Nigel M. Allinson (Univ. Manchester Institute of Technology - UK) Jose Nelson Amaral (Pontificia Catholic University, RS - BRAZIL) William Armstrong (University of Alberta - CANADA) Valeriu Beiu (Los Alamos National Laboratory - USA) Dibio Leandro Borges * (Federal University of Goias - BRAZIL) Antonio de Padua Braga (Federal University of Minas Gerais - BRAZIL) Andre L.P. de Carvalho (University of Sao Paulo, S. Carlos - BRAZIL) Weber Martins * (Federal University of Goias - BRAZIL) Edilberto P. Teixeira (Federal University of Uberlandia - BRAZIL) Harold Szu (University of Southwestern Louisiana - USA) Germano C. Vasconcelos (Federal University of Pernambuco - BRAZIL) * Program Chairmen T i m e t a b l e ----------------------------------------------------------------------- Submissions due May 2nd, 1997 Notification of acceptance July 5th, 1997 Final manuscript due August 5th, 1997 Conference date December 3rd-5th, 1997 P a p e r S u b m i s s i o n s ----------------------------------------------------------------------- We invite submissions of scientific papers to any topic related to Neural Networks and Evolutionary Computation. Authors should submit four (4) copies of their papers in hard copy form. Neither computer files nor fax submission are acceptable. Submissions must be printed on 8 1/2 x 11 inch (21.59 x 27.94 cm) or A4 paper using 12 point type, and they must be a maximum of twenty (20) pages long, double space, including all figures and references. The reviewing process will be blind to the identities of the authors, please note that this requires that authors exercise some care not to identify themselves in their papers. Title page and paper body ------------------------- Each copy of the paper must include a title page, separate from the body of the paper, containing the title of the paper, the names and addresses of all authors (please, complete addresses including also e-mail and FAX), a short abstract of less than 200 words, and a list of keywords giving the area/subarea of the paper. The second page, on which the paper body begins, should include the same title, abstract, keywords, but not the names and affiliations of the authors. All paper submissions should be to the following address: --------------------------------------------------------- IVth Brazilian Symposium on Neural Networks (SBRN'97) (PAPER SUBMISSION) School of Electrical Engineering (EEE) Federal University of Goias (UFG) Pca. Universitaria s/n Setor Universitario 74605-220 Goiania, GO (BRAZIL) FAX: +55 62 202 - 0325 E-mail: sbrn97 at eee.ufg.br Review criteria --------------- Papers will be subject to peer review by an international program committee. Selection criteria include accuracy and originality of ideas, clarity and significance of results and the quality of the presentation. P u b l i c a t i o n ----------------------------------------------------------------------- We are planning to publish the international proceedings with a major publishing house, and details will be available soon. Papers submitted in Portuguese or Spanish will be published in a separate volume. Both proceedings will be available at the conference. Please, note that at least one author should attend the conference for presentation of the accepted paper. I n q u i r i e s ----------------------------------------------------------------------- Inquiries regarding any aspect of the conference may be sent to the internet address: sbrn97 at eee.ufg.br or to the Conference Chairmen: Weber Martins and Dibio Leandro Borges both at: School of Electrical Engineering Federal University of Goias Pca. Universitaria s/n Setor Universitario 74605-220 Goiania, GO (BRAZIL) \/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\ From jbednar at cs.utexas.edu Fri Feb 14 17:11:31 1997 From: jbednar at cs.utexas.edu (jbednar@cs.utexas.edu) Date: Fri, 14 Feb 1997 16:11:31 -0600 Subject: Overview paper and theses on the RF-LISSOM project Message-ID: <199702142211.QAA35148@duckula.cs.utexas.edu> The following paper (to appear in Psychology of Learning and Motivation) gives an overview of the RF-LISSOM project on modeling the primary visual cortex, ongoing at the UTCS Neural Networks Research Group. For those seeking details, please refer to the dissertation and the theses also announced below. All these publications (and others) are available from our web page at http://www.cs.utexas.edu/users/nn (under publications, self-organization; the direct ftp addresses are included below as well). Public domain LISSOM code, for self-organization in laterally connected maps, will be announced in the near future. -- The Authors ------------------------------------------------------------------------- SELF-ORGANIZATION, PLASTICITY, AND LOW-LEVEL VISUAL PHENOMENA IN A LATERALLY CONNECTED MAP MODEL OF THE PRIMARY VISUAL CORTEX Risto Miikkulainen, James A. Bednar, Yoonsuck Choe, and Joseph Sirosh Department of Computer Sciences, The University of Texas at Austin. In R. L. Goldstone, P. G. Schyns, and D. L. Medin (eds.), Psychology of Learning and Motivation, vol. 36, 1997 in press (36 pages). ftp://ftp.cs.utexas.edu/pub/neural-nets/papers/miikkulainen.visual-cortex.ps.Z Based on a Hebbian adaptation process, the afferent and lateral connections in the RF-LISSOM model organize simultaneously and cooperatively, and form structures such as those observed in the primary visual cortex. The neurons in the model develop local receptive fields that are organized into orientation, ocular dominance, and size selectivity columns. At the same time, patterned lateral connections form between neurons that follow the receptive field organization. This structure is in a continuously-adapting dynamic equilibrium with the external and intrinsic input, and can account for reorganization of the adult cortex following retinal and cortical lesions. The same learning processes may be responsible for a number of low-level functional phenomena such as tilt aftereffects, and combined with the leaky integrator model of the spiking neuron, for segmentation and binding. The model can also be used to verify quantitatively the hypothesis that the visual cortex forms a sparse, redundancy-reduced encoding of the input, which allows it to process massive amounts of visual information efficiently. ------------------------------------------------------------------------- A SELF-ORGANIZING NEURAL NETWORK MODEL OF THE PRIMARY VISUAL CORTEX Joseph Sirosh Department of Computer Sciences, The University of Texas at Austin. PhD Dissertation and Technical Report AI95-237, August 1995 (137 pages). ftp://ftp.cs.utexas.edu/pub/neural-nets/papers/sirosh.diss.tar This work is aimed at modeling and analyzing the computational processes by which sensory information is learned and represented in the brain. First, a general self-organizing neural network architecture that forms efficient representations of visual inputs is presented. Two kinds of visual knowledge are stored in the cortical network: information about the principal feature dimensions of the visual world (such as line orientation and ocularity) is stored in the afferent connections, and correlations between these features in the lateral connections. During visual processing, the cortical network filters out these correlations, generating a redundancy-reduced sparse coding of the visual input. Through massively parallel computational simulations, this architecture is shown to give rise to structures similar to those in the primary visual cortex, such as (1) receptive fields, (2) topographic maps, (3) ocular dominance, orientation and size preference columns, and (4) patterned lateral connections between neurons. The same computational process is shown to account for many of the dynamic processes in the visual cortex, such as reorganization following retinal and cortical lesions, and perceptual shifts following dynamic receptive field changes. These results suggest that a single self-organizing process underlies development, plasticity and visual functions in the primary visual cortex. ------------------------------------------------------------------------- TILT AFTEREFFECTS IN A SELF-ORGANIZING MODEL OF THE PRIMARY VISUAL CORTEX James A. Bednar Department of Computer Sciences, The University of Texas at Austin. Masters Thesis and Technical Report AI97-259, January 1997 (104 pages). ftp://ftp.cs.utexas.edu/pub/neural-nets/papers/bednar.thesis.tar The psychological phenomenon known as the tilt aftereffect was used to demonstrate the functional properties of RF-LISSOM, a self-organizing model of laterally connected orientation maps in the primary visual cortex. The same self-organizing processes that are responsible for the development of the map and its lateral connections are shown to result in tilt aftereffects as well. The model allows analysis of data that are difficult to measure in humans, thus providing a view of the cortex that is otherwise not available. The results give computational support for the idea that tilt aftereffects arise from lateral interactions between adapting feature detectors, as has long been surmised. They also suggest that indirect tilt aftereffects could result from the conservation of synaptic resources. The model thus provides a unified computational explanation of self-organization and both direct and indirect tilt aftereffects in the primary visual cortex. ------------------------------------------------------------------------- LATERALLY INTERCONNECTED SELF-ORGANIZING FEATURE MAP IN HANDWRITTEN DIGIT RECOGNITION Yoonsuck Choe Department of Computer Sciences, The University of Texas at Austin. Masters Thesis and Technical Report AI95-236, August 1995 (65 pages). ftp://ftp.cs.utexas.edu/pub/neural-nets/papers/choe.thesis.tar.Z An application of biologically motivated laterally interconnected synergetically self-organizing maps (LISSOM) to off-line recognition of handwritten digit is presented. The lateral connections of the LISSOM map learn the correlations between units through Hebbian learning. As a result, the excitatory connections focus the activity in local patches and lateral connections decorrelate redundant activity on the map. This process forms internal representations for the input that are easier to recognize than the input bitmaps themselves or the activation patterns on a standard Self-Organizing Map (SOM). The recognition rate on a publically available subset of NIST special database 3 with LISSOM is 4.0% higher than that based on SOM, and 15.8% higher than that based on raw input bitmaps. These results form a promising starting point for building pattern recognition systems with a LISSOM map as a front end. From harnad at cogsci.soton.ac.uk Sun Feb 16 09:44:27 1997 From: harnad at cogsci.soton.ac.uk (Stevan Harnad) Date: Sun, 16 Feb 97 14:44:27 GMT Subject: Feature Creation: BBS Call for Commentators Message-ID: <20512.9702161444@cogsci.ecs.soton.ac.uk> Below is the abstract of a forthcoming BBS target article on: THE DEVELOPMENT OF FEATURES IN OBJECT CONCEPTS by Philippe G. Schyns, Robert L. Goldstone & Jean-Pierre Thibaut This article has been accepted for publication in Behavioral and Brain Sciences (BBS), an international, interdisciplinary journal providing Open Peer Commentary on important and controversial current research in the biobehavioral and cognitive sciences. Commentators must be BBS Associates or nominated by a BBS Associate. To be considered as a commentator for this article, to suggest other appropriate commentators, or for information about how to become a BBS Associate, please send EMAIL to: bbs at cogsci.soton.ac.uk or write to: Behavioral and Brain Sciences Department of Psychology University of Southampton Highfield, Southampton SO17 1BJ UNITED KINGDOM http://www.princeton.edu/~harnad/bbs/ http://www.cogsci.soton.ac.uk/bbs/ ftp://ftp.princeton.edu/pub/harnad/BBS/ ftp://ftp.cogsci.soton.ac.uk/pub/bbs/ gopher://gopher.princeton.edu:70/11/.libraries/.pujournals If you are not a BBS Associate, please send your CV and the name of a BBS Associate (there are currently over 10,000 worldwide) who is familiar with your work. All past BBS authors, referees and commentators are eligible to become BBS Associates. To help us put together a balanced list of commentators, please give some indication of the aspects of the topic on which you would bring your areas of expertise to bear if you were selected as a commentator. An electronic draft of the full text is available for inspection with a WWW browser, anonymous ftp or gopher according to the instructions that follow after the abstract. ____________________________________________________________________ THE DEVELOPMENT OF FEATURES IN OBJECT CONCEPTS Philippe G. Schyns Robert L. Goldstone Jean-Pierre Thibaut Psychology Dept. Psychology Dept. Psychology Dept. Glasgow University Indiana University Universite de Liege Glasgow G12 8QB UK Bloomington IN 47405 4000 Liege BELGIUM philippe at psy.gla.ac.uk rgoldsto at ucs.indiana.edu jthibaut at vm1.ulg.ac.be KEYWORDS: Concept learning, conceptual development, perceptual learning, features, stimulus encoding ABSTRACT: According to an influential approach to cognition, our perceptual systems provide us with a repertoire of fixed features as input to higher-level cognitive processes. We present a theory of category learning and representation in which features, instead of being components of a fixed repertoire, are created under the influence of higher-level cognitive processes. When new categories need to be learned, fixed features face one of two problems: (1) High-level features that are directly useful for categorization may not be flexible enough to represent all relevant objects. (2) Low-level features consisting of unstructured fragments (such as pixels) may not capture the regularities required for successful categorization. We report evidence that feature creation occurs in category learning and we describe the conditions that promote it. Feature creation can adapt flexibly to changing environmental demands and may be the origin of fixed feature repertoires. Implications for object categorization, conceptual development, chunking, constructive induction and formal models of dimensionality reduction are discussed. -------------------------------------------------------------- To help you decide whether you would be an appropriate commentator for this article, an electronic draft is retrievable from the World Wide Web or by anonymous ftp or gopher from the US or UK BBS Archive. Ftp instructions follow below. Please do not prepare a commentary on this draft. Just let us know, after having inspected it, what relevant expertise you feel you would bring to bear on what aspect of the article. The URLs you can use to get to the BBS Archive: http://www.princeton.edu/~harnad/bbs/ http://www.cogsci.soton.ac.uk/bbs/Archive/bbs.schyns.html ftp://ftp.princeton.edu/pub/harnad/BBS/bbs.schyns ftp://ftp.cogsci.soton.ac.uk/pub/bbs/Archive/bbs.schyns gopher://gopher.princeton.edu:70/11/.libraries/.pujournals To retrieve a file by ftp from an Internet site, type either: ftp ftp.princeton.edu or ftp 128.112.128.1 When you are asked for your login, type: anonymous Enter password as queried (your password is your actual userid: yourlogin at yourhost.whatever.whatever - be sure to include the "@") cd /pub/harnad/BBS To show the available files, type: ls Next, retrieve the file you want with (for example): get bbs.schyns When you have the file(s) you want, type: quit From becker at curie.psychology.mcmaster.ca Sun Feb 16 20:58:15 1997 From: becker at curie.psychology.mcmaster.ca (Sue Becker) Date: Sun, 16 Feb 1997 20:58:15 -0500 (EST) Subject: COMPUTATIONAL NEUROSCIENCE POSTDOCTORAL POSITION AVAILABLE Message-ID: COMPUTATIONAL NEUROSCIENCE POSTDOCTORAL POSITION AVAILABLE Department of Psychology McMaster University A postdoctoral position is open in the Psychology Department of McMaster University, Hamilton, Ontario. A multidisciplinary approach will be taken to develop biologically plausible models of hippocampal and neocortical memory systems. Projects will include developing simulations of cortical-cortical and subcortical-cortical interactions during learning and information storage. In addition to our focus on processing in hippocampal-cortical systems, we are also investigating and modelling the role of cortico-thalamic back-projections. Research projects will be conducted in close collaboration with R. Racine, a neuroscientist, S. Becker, a computational modeller, S. Haykin, an electrical engineer, and G. Gerstein, a computational neuroscientist. The primary goal of this collaborative effort is to build powerful learning algorithms in neural networks which are based on rules suggested by both memory research and physiology research (e.g. LTP work). Racine's laboratory has recently provided the first demonstrations of LTP in the neocortex and cortico-thalamic connections in chronic preparations. The rules that apply to LTP induction in these systems are quite different from those determined for the hippocampus. An optional component of this postdoctoral position would be participation in further experimental investigations of neocortical LTP in either slice or in vivo preparations. This project is funded by a collaborative research grant from the Natural Sciences and Engineering Research Council of Canada to R. Racine, G. Gerstein, S. Haykin and S. Becker. Please send curriculum vitae, expression of interest, and the names and e-mail or phone numbers of three references to Ron Racine at racine at mcmail.cis.mcmaster.ca From Gerhard.Paass at gmd.de Mon Feb 17 05:48:15 1997 From: Gerhard.Paass at gmd.de (Gerhard Paass) Date: Mon, 17 Feb 1997 11:48:15 +0100 Subject: CfP: Interdisciplinary College IK'97 Message-ID: <199702171048.AA02418@sein.gmd.de> Call for Participation Interdisciplinary College IK'97 This conference aims at establishing a link between four subjects: biology of neurons, neural information science, artificial intelligence and cognitive psychology. Leading scientists of each field provide introductory courses on each subject and give an in-depth account of current developments. In addition six interdisciplinary courses between each of the areas are offered, where two scientists of each field give an integrative view. Conference Language: German Time: March 15.-22.th 1997, Location: Guenne/Moehnesee, close to Dortmund/Germany Deadline for Registration: Feb. 28th (payment of conference fees) WWW: http://www.informatik.uni-ulm.de/ni/ik97.html From georg at ai.univie.ac.at Mon Feb 17 12:06:52 1997 From: georg at ai.univie.ac.at (Georg Dorffner) Date: Mon, 17 Feb 1997 18:06:52 +0100 (MET) Subject: Open position: NN for time series processing Message-ID: <199702171706.SAA21667@jedlesee.ai.univie.ac.at> The Neural Networks Group at the Austrian Research Institute for Artificial Intelligence in Vienna, Austria, has an immediate opening for a position on Neural Networks for Time Series Processing and Economic Modeling as part of the larger initiative "Adaptive Models in Economics and Management Science" (http://www.wu-wien.ac.at/am/am.html). The goal is to investigate the capabilities of recurrent neural nets in modeling economic processes, including their use in multi-agent simulations, with a focus on learning algorithms, adaptivity, and mixture-of-expert approaches. Required background: - knowledge in practical time series processing (e.g. ARMA, state space models, etc.) - basic knowledge in neural networks and programming - Ph.D. degree or comparable doctorate - research experience Desired background: - publications in the field - basic knowledge in economic modeling - basic knowledge in statistics - experience in a UNIX environment - basic understanding knowledge of German The position will be available for up to three years, with a monthly salary of about 18.000 ATS (approx. 1500 US$) per month, 14 times a year, after tax. Interested applicants should apply by email, mail or fax, no later than March 5, 1997 at the address below. Include a short vita, a list of publications, and any other information demonstrating your qualification. Send applications to: Georg Dorffner Austrian Research Institute for Artificial Intelligence Schottengasse 3 A-1010 Vienna phone: +43-1-53532810 fax: +43-1-5320652 email: georg at ai.univie.ac.at This position is funded by the Austrian Fonds zur Foerderung der wissenschaftlichen Forschung. From reggia at cs.umd.edu Mon Feb 17 15:11:29 1997 From: reggia at cs.umd.edu (James A. Reggia) Date: Mon, 17 Feb 1997 15:11:29 -0500 (EST) Subject: New Book: Neural Modeling of Brain and Cognitive Disorders Message-ID: <199702172011.PAA16491@avion.cs.umd.edu> NEURAL MODELING OF BRAIN AND COGNITIVE DISORDERS edited by James A Reggia (Univ. Maryland), Eytan Ruppin (Tel Aviv Univ.) & Rita Sloan Berndt (Univ. Maryland) During the last few years there has been a rapidly increasing interest in neural modeling of brain and cognitive disorders. This multidisciplinary book presents a variety of such models in neurology, neuropsychology and psychiatry. A review of work in this area is given first. Computational models are then presented of memory impairment in Alzheimer's disease, functional brain reorganization following a stroke, patterns of neural activity in epilepsy, disruption of language processes in aphasia and acquired dyslexia, altered cognitive processes in schizophrenia and depression, and related disorders. 425pp (approx.) 1996 981-02-2879-1 US$86 Ordering Information: Please e-mail or fax your order and VISA/MC/AMEX/Diners Club Card Number or alternatively enclose a cheque/bank draft and send to your nearest World Scientific Office: USA: World Scientific Publishing Co. Inc. 1060 Main Street, River Edge, NJ 07661, USA Toll-free: 1-800-227-7562 Toll-free fax: 1-888-977-2665 E-mail: wspc at wspc.com -UK: World Scientific Publishing Co. Ltd. 57 Shelton Street, Covent Garden, London WC2H 9HE, UK Tel: 44-171-836-0888 Fax: 44-171-836-2020 E-mail: sales at wspc2.demon.co.uk -Singapore: World Scientific Publishing Co. Ltd. Farrer Road P.O. Box 128, Singapore 912805 Tel: 65-382-5663 Fax: 65-382-5919 From zhangw at redwood.rt.cs.boeing.com Mon Feb 17 19:48:02 1997 From: zhangw at redwood.rt.cs.boeing.com (Wei Zhang) Date: Mon, 17 Feb 1997 16:48:02 -0800 Subject: Job: machine learning at Boeing Message-ID: <199702180048.QAA04330@rivendell.network-A> **Outstanding Machine Learning Researcher needed** The Boeing Company, the world's largest aerospace company, is actively working research projects in advanced computing technologies including projects involving NASA, FAA, Air Traffic Control, and Global Positioning as well as airplane and manufacturing research. The Research and Technology organization located in Bellevue, Washington, near Seattle, has an open position for a machine learning researcher. We are the primary computing research organization for Boeing and have contributed heavily to both short term technology advances and to long range planning and development. BACKGROUND REQUIRED: Machine Learning, Knowledge Discovery, Data Mining, Statistics, Artificial Intelligence or related field. RESEARCH AREAS: We are developing and applying techniques for data mining and statistical analyses of diverse types of data, including: safety incidents, flight data recorders, reliability, maintenance, manufacturing, and quality assurance data. These are not areas where most large R&D data mining efforts are currently focused. Research areas include data models, data mining algorithms, statistics, and visualization. Issues related to our projects also include pattern recognition, multidimensional time series, and temporal databases. We can achieve major practical impacts in the short-term both at Boeing and in the airline industry, which may result in a safer and more cost-effective air travel industry. A Ph.D. in Computer Science or equivalent experience is highly desirable for the position. We strongly encourage diversity in backgrounds including both academic and industrial experiences. Knowledge of machine learning, statistics, and data mining are important factors. Experience with databases and programming (C/C++, JAVA, and Splus) is desirable. APPLICATION: If you meet the requirements and you are interested, please send your resume via electronic e-mail in plain ASCII format to zhangw at redwood.rt.cs.boeing.com (Wei Zhang). You can also send it via US mail to Wei Zhang The Boeing Company PO Box 3707, MS 7L-66 Seattle, WA 98124-2207 Application deadline is April 30, 1997. The Boeing Company is an equal opportunity employer. From parodi at cptsu2.univ-mrs.fr Tue Feb 18 07:54:31 1997 From: parodi at cptsu2.univ-mrs.fr (Olivier Parodi) Date: Tue, 18 Feb 97 13:54:31 +0100 Subject: Cargese Summer School on Neural Information Processing Message-ID: <9702181254.AA19906@cptsu2.univ-mrs.fr> SUMMER SCHOOL on NEURONAL INFORMATION PROCESSING : FROM BIOLOGICAL DATA TO MODELLING AND APPLICATIONS Cargese -- Corse du Sud (France) June 30 -- July 12, 1997 organized at the INSTITUT D'ETUDES SCIENTIFIQUES DE CARGESE CNRS (UMS 820) Universite de Nice-Sophia Antipolis - Universite de Corte F 20130 CARGESE Sponsored by Centre National de la Recherche Scientifique DGA-DRET (French Ministry of Defence) Conseil Executif de la Corse THEME In the past ten years, statistical mechanics and dynamics of neuronal automata have been extensively studied. Most of the work has been based on over-simplified models of neurons. Recent developments in Neurosciences have, however, considerably modified our knowledge of both the operating modes of neurons and information processing in the cortex. Multi-unit recordings have allowed precise temporal correlations to be detected, within temporal windows of the order of 1 ms. Simultaneously, oscillations corresponding to a quasi-periodic spike-firing, synchronized over several visual cortical areas, have been observed with anaesthesied cats and with monkeys. Last but not least, recent work on the neuronal operating modes have emphasized the role played by the dendritic arborization. These developments have led to considerable interest for coding scheme relying on precise spatio-temporal patterns both from the theoretical and experimental points of view. This prompts us to consider, for information processing, new models which would proceed, e.g., from a synchronous detection of correlated spike firing, and could be particularly robust against noise. Such models might bring about original technical applications for information processing and control. Further developments in this field may be of major importance for our understanding of the basic mechanisms of perception and cognition. They should also lead to new concepts in applications directed towards artificial perception and pattern recognition. Up to now, artificial systems for pattern recognition are far from reaching the standards of human vision. Systems based on a temporal coding by spikes may now be expected to bring about major improvements in this field. The aim of the school is to provide students and people engaged in both applied and basic research with state of the art in every relevant field (Neurosciences, Physics, Mathematics, Information and Control Theory) and to encourage further interdisciplinary and international exchanges. PRELIMINARY PROGRAM Pr. M. Abeles (Jerusalem) - Spike Correlations and synfire chains, Pr. P. Combe (Marseille) - Learning: a geometrical approach, Pr. J. Demongeot (Grenoble) - Dynamics of modular architectures, Pr. L. van Hemmen (Munchen) - Coding by spikes, Pr. J. Herault (Grenoble) - Information processing in the retina, Pr. J. Hertz (Copenhague) - Temporal coding and learning, Pr. A. Holley (Lyon) - Information processing in the olfactory mammalian system, Pr. C. von der Malsburg (Bochum) - Binding and Synchrony, Dr. C. Masson (Paris) - Cellular mechanism of odor discrimination, Dr. J. P. Nadal (Paris) - Information theory and neuronal architecture, Dr. O. Parodi (Marseille) - Temporal coding and correlation detection, Pr. S. Solla (ATT) - Dynamics of on-line learning processes. OFFICIAL LANGUAGES The official languages of the School are English and French. Lectures will given in English. DIRECTOR OF THE SCHOOL Olivier PARODI, Centre de Physique Theorique CNRS-Luminy, case 907, F 13288 MARSEILLE CEDEX 09, France Tel. (33) 4 91 26 95 30 Fax (33) 4 91 26 95 53 e-mail parodi at cpt.univ-mrs.fr REGISTRATION FEES Students: free CNRS and members of CNRS institutes: free University: 1500 FF Industry: 2500 FF ACCOMMODATION GRANTS 1 - The School is sponsored by the Formation Permanente du CNRS, which can support accomodation expenses of at least 16 CNRS participants. 2 - The Organizing Committee will consider grants for students and young participants. PRACTICAL INFORMATION The school will be held at the Institut d'Etudes Scientifiques de Cargese. Lectures and Seminars will be given from 9 am to 12:30 pm and from 4 to 7:15 pm.(except on Sunday) from July 1 to 11 included. All participants are expected to arrive on June 30 and leave on July 12. TRAVEL : Cargese is located approximatively 60 kms North of Ajaccio. The best way to get to Cargese is to reach Ajaccio. You are asked to make your own travelling arrangements. However, in order to reduce the cost of your travel, two groups will be organized on regular flights between Paris and Ajaccio and Marseille and Ajaccio on June 30 and July 12 (approx. 1200-1400 FF for a Paris- Ajaccio return ticket). We also consider renting buses for the Ajaccio-Cargese and return journeys. Details will be sent later. ACCOMODATION : The Institute is located 2 km south of the village, on the sea shore. On working days, lunches will be served at the Institute (800 FF for the session, including refreshment and coffe breaks). There are various housing possibilities: - shared room at the Institute or in apartments in the village - 1920 FF (per person for the session) - single room in shared apartments in the village - 2640 FF for the session - rented apartment in the village for your family - 270 to 500 FF per day - hotels in the village - 250 to 400 FF per person per day - camping on the grounds of the Institute - 20 FF per person per day. You have to bring your own equipment, showers are at your disposal. Breakfast will be served at the Institute for people staying on the grounds. WARNING: There are no banks nor cash machines in Cargese and credit cards are not accepted everywhere. Think to bring enough French cash before leaving Paris, Marseille or Ajaccio. NOTE : We cannot provide accomodation before or after the dates of the School. POSTER SESSION : One or several poster sessions will be organized. Participants are encouraged to prepare posters on their own work. REGISTRATION : You have to fill the enclosed application and return it by e-mail to cargese at cpt.univ-mrs.fr BEFORE MARCH 31, 1997 with, if necessary, a letter justifying your grant request. The School will accept up to 50 students, including those supported by the Formation Permanente of CNRS. In case of over-demand, participants will be selected by the Scientific Committee, with a balance between junior and senior students, and a preference for students carrying an active research in the field. ------------------ http://cpt.univ-mrs.fr/Cargese ---------------------- ***************************************************************************** ***************************************************************************** French Version Version Francaise (TEX) a compiler en "plain TEX" sur une station UNIX, un MAC avec Textures ou un PC. ***************************************************************************** \nopagenumbers \vsize 26.5 true cm \hsize 18.5 true cm \hoffset -1.2 true cm \voffset -1 cm \null {\bf \centerline{ECOLE D'ETE} \vskip 1 true cm \centerline{TRAITEMENT NEURONAL DE L'INFORMATION} \medskip \centerline{Des donn\'ees biologiques \`a la mod\'elisation et aux applications} \vskip 1 true cm \centerline{Carg\`ese, Corse du Sud (France)} \par \centerline{30 juin -- 12 juillet 1997} \bigskip \centerline{organis\'ee \`a} \bigskip \centerline{L'INSTITUT D'ETUDES SCIENTIFIQUES DE CARGESE} } \smallskip \centerline{CNRS (UMS 820) -- Universit\'e de Nice-Sophia Antipolis -- Universit\'e de Corte} \smallskip \centerline{F 20130 Carg\`ese} \vskip 1 true cm \noindent Subventionn\'ee par \par Le Centre National de la Recherche Scientifique \par La Direction Scientifique de la DGA-DRET (Minist\`ere de la D\'efense) \par La Collectivit\'e Territoriale Corse \vskip 1 true cm\noindent {\bf THEME} \medskip Depuis une dizaine d'ann\'ees, de nombreuses \'etudes ont port\'e sur la m\'ecanique statistique et la dynamique d'au\-to\-ma\-tes neuronaux, bas\'ees le plus souvent sur une mod\'elisation relativement ancienne du neurone. Mais les d\'eveloppements r\'ecents en Neurosciences ont consid\'erablement modifi\'e nos connaissances, \`a la fois sur le neurone et sur le traitement de l'information dans le cortex. \par Les techniques d'enregistrement multi-\'electrodes ont permis de d\'etecter des corr\'elations temporelles tr\`es pr\'ecises dans des fen\^etres de l'ordre de la milliseconde. En m\^eme temps des {\it oscillations}, correspondant \`a des \'emissions quasi-p\'eriodiques de potentiels d'action synchronis\'ees \`a travers tout le cortex visuel, ont \'et\'e observ\'ees chez des chats anesth\'esi\'es et chez des singes. Parall\`element, des travaux portant sur le fonctionnement du neurone ont soulign\'e le r\^ole important jou\'e par l'arborisation dendritique, r\^ole compl\`etement pass\'e sous silence dans les mod\`eles classiques d'automates neuronaux. \par Ces d\'eveloppements ont donn\'e un essor consid\'erable \`a des \'etudes, tant th\'eoriques qu'exp\'erimentales, faisant intervenir un codage par des configurations spatio-temporelles pr\'ecises. On est ainsi amen\'e \`a envisager de nouveaux mod\`eles de traitement de l'information, bas\'es, par exemple, sur la d\'etection synchrone de signaux corr\'el\'es, et qui pourraient \^etre particuli\`erement robustes vis-\`a-vis du bruit. Ces mod\`eles pourraient conduire \`a des applications technologiques originales en mati\`ere de traitement de l'information et de contr\^ole. \par Le but de cette \'ecole est d'apporter aux \'etudiants et aux chercheurs, tant fondamentaux qu'ap\-pli\-qu\'es, un enseignement de pointe couvrant, dans ce domaine, l'ensemble des disciplines concern\'ees (Neurosciences, Physique, Math\'ematiques, Th\'eories de l'Information et du Contr\^ole), et de favoriser des coop\'erations interdisciplinaires et internationales. \vskip .8 true cm\noindent {\bf PROGRAMME PRELIMINAIRE} \medskip \par Pr. M. Abeles (Jerusalem) -- Spike Correlations and synfire chains, \par Pr. P. Combe (Marseille) -- Learning: a geometrical approach, \par Pr. J. Demongeot (Grenoble) -- Dynamics of modular architectures, \par Pr. L. van Hemmen (Munchen) -- Coding by spikes, \par Pr. J. H\'erault (Grenoble) -- Information processing in retina, \par Pr. J. Hertz (Copenhague) -- Temporal coding and learning, \par Pr. A. Holley (Lyon) -- Information processing in the olfactory mammalian system, \par Pr. C. von der Malsburg (Bochum) -- Binding and Synchrony, \par Dr. C. Masson (Paris) -- Cellular mechanism of odor discrimination, \par Dr. J. P. Nadal (Paris) -- Information theory and neuronal architecture, \par Dr. O. Parodi (Marseille) -- Temporal coding and correlation detection, \par Pr. S. Solla (ATT) -- Dynamics of on-line learning processes \vfill \eject \noindent {\bf LANGUES OFFICIELLES} \smallskip Les langues officielles de l'Ecole sont l`anglais et le fran\c cais. Les cours seront donn\'es en anglais \bigskip\noindent {\bf DIRECTEUR DE L'ECOLE} \smallskip Olivier PARODI, Centre de Physique Th\'eorique \par CNRS-Luminy, case 907, F 13288 MARSEILLE CEDEX 09, France \par Tel. (33) 4 91 26 95 30 Fax (33) 4 91 26 95 53 \par e-mail parodi at cpt.univ-mrs.fr \bigskip\noindent {\bf DROITS D'INSCRIPTION} \smallskip \par Etudiants: n\'eant \par Personnels CNRS et membres de laboratoires soutenus par le CNRS: n\'eant \par Autres universitaires: 1500 FF \par Secteur industriel: 2500 FF \bigskip\noindent {\bf ALLOCATIONS} \smallskip \item{1 -} L'Ecole est subventionn\'ee par la Formation permanente du CNRS. Les frais de s\'ejour d'au moins 16 participants CNRS -- ou, \`a d\'efaut, de participants en provenance de Laboratoires soutenus par le CNRS -- pourront \^etre pris en charge par la Formation Permanente de la XIIe D\'el\'egation. Les frais de transport des personnels CNRS seront alors pris en charge par la Formation Permanente de leur D\'el\'egation. \item{2 -} Le Comit\'e d'Organisation envisage d'attribuer des bourses aux jeunes chercheurs. \bigskip\noindent {\bf INFORMATIONS PRATIQUES} \smallskip L'Ecole est organis\'ee \`a l'Institut d'Etudes Scientifiques de Carg\`ese. Les Cours et les S\'eminaires auront lieu de 9h \`a 12h30 et de 16h \`a 19h15 tous les jours, sauf le Dimanche, du 1er juillet au 11 juillet inclus. Tous les participants sont attendus le 30 juin et leur d\'epart est pr\'evu le 12 juillet. \medskip\noindent VOYAGE: \par Carg\`ese est situ\'e \`a environ 60 km au Nord d'Ajaccio. La meilleure mani\`ere de s'y rendre est de gagner d'abord Ajaccio. Vous \^etes responsable de l'organisation de votre voyage, aller et retour. Cependant, afin d'en r\'eduire le co\^ut, des vols de groupe sur ligne r\'eguli\`ere seront organis\'es entre Paris et Ajaccio et Marseille et Ajaccio les 30 juin et 12 juillet (Prix approximatif 1200 \`a 1400 FF pour un aller et retour Paris-Ajaccio). Nous envisageons aussi de louer des cars pour les aller et retour Ajaccio-Carg\`ese. Les d\'etails seront communiqu\'es ult\'erieurement. \medskip\noindent LOGEMENT: \par L'Institut est situ\'e \`a 2 km du village au bord de la mer. Les jours de cours, le d\'ejeuner est servi sur place (800 FF pour la session y compris les pauses-caf\'e). Les possibilit\'es d'h\'ebergement sont diverses: \medskip \item{- } Chambre partag\'ee avec un coll\`egue \`a l'Institut ou dans un appartement au village (1920 FF pour la dur\'ee de la session), \item{- } Chambre individuelle dans un appartement partag\'e (2640 FF pour la session), \item{- } Appartement au village en location pour votre famille (270 \`a 500 FF par jour) \item{- } H\^otel dans le village (250 \`a 400 FF par jour et par personne) \item{- } Camping sur le terrain m\^eme de l'Institut, avec votre mat\'eriel individuel (20 FF par jour). \medskip\noindent {\it ATTENTION: Pas de banque ni de distributeur automatique de billet \`a Carg\`ese, et les cartes de cr\'edit ne sont pas toujours accept\'ees. Pensez \`a vous munir d'argent liquide.} \medskip\noindent NB: Nous n'organisons aucun h\'ebergement en dehors des dates de l'Ecole. \bigskip\noindent {\bf COMMUNICATIONS AFFICH\'EES} \par Une ou plusieurs s\'eances de communications par affiches seront organis\'ees. Les participants sont invit\'es \`a pr\'eparer des affiches sur leur propre travail de recherche. \bigskip\noindent {\bf INSCRIPTION} \par Le formulaire ci-joint doit \^etre renvoy\'e au plus tard le {\bf 31 mars 1997} par courrier \'electronique \`a \break {\bf cargese at cpt.univ-mrs.fr}, accompagn\'e, \'eventuellement, d'une lettre justifiant votre demande de bourse ou de prise en charge par la Formation Permanente du CNRS. \par Le nombre de participants \`a l'Ecole, enseignants non compris, est limit\'e \`a cinquante, nombre incluant les stagiaires subventionn\'es par la Formation Permanente du CNRS. La s\'election des candidats sera effectu\'ee par le Comit\'e Scientifique, qui veillera \`a l'\'equilibre entre jeunes chercheurs et chercheurs confirm\'es, et donnera une priorit\'e aux participants engag\'es dans une recherche active dans le domaine. \medskip \noindent {\bf WEB} \par Toutes les informations concernant l'Ecole sont disponibles \`a l'URL \par {\bf http://cpt.univ-mrs.fr/Cargese/} \end *************************************************************************** *************************************************************************** ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||| REGISTRATION FORM BULLETIN D'INSCRIPTION NEURONAL INFORMATION PROCESSING TRAITEMENT NEURONAL DE L'INFORMATION NAME / NOM First name / Prenom (Ms/Mr) Nationality/Nationalite date of birth / date INSTITUTE ADDRESS / LABORATOIRE Tel. Fax e-mail HOME ADDRESS / Adresse personnelle Tel. Next of kin / Personne a prevenir en cas d'accident : DIPLOMA /Niveau Universitaire PRESENT POSITION / EMPLOI ACTUEL ? PRESENT RESEARCH / RECHERCHES ACTUELLES MAIN PUBLICATIONS / Publications Principales (up to five recent publications) Will your Institution pay for / Votre Laboratoire subventionne-t-il: your travel (Yes/No) ? votre voyage (Oui/Non) ? your accommodation expenses (Yes/No) ? vos frais de sejour ? (Oui/Non) ? Only for members of the CNRS / Pour les participants CNRS, ou relevant d'un laboratoire soutenu par le CNRS, demandez-vous la prise en charge de vos frais de sejour et de transport par la Formation Permanente du CNRS ? For Junior applicants, do you request a grant ? (Yes/No) ? Pour les jeunes chercheurs, etes-vous candidat a une bourse ? (Oui/Non) RECOMMENDATION : Junior applicants should arrange for a letter of recommendation from a senior scientist who had the oppotunity to judge their work. This letter should be sent by e-mail to cargese at cpt.univ-mrs.fr. RECOMMANDATION: Les jeunes chercheurs doivent fournir une lettre de recommandation d'un chercheur confirme. Cette lettre doit etre envoyee par courrier electronique a cargese at cpt.univ-mrs.fr YOU ARE RECOMMENDED BY / Vous etes recommande par ? TRAVEL/VOYAGE: Will you take a group flight / Prendrez-vous un vol de groupe ? Paris-Ajaccio-Paris : Number/Nombre ? Marseille-Ajaccio-Marseille : Number/Nombre ? HOUSING (Specify your choice): LOGEMENT (Precisez votre choix): Pour les participants francophones: Etes-vous capable de suivre un cours en anglais ? De participer a une discussion en anglais ? DEADLINE / DATE LIMITE D'INSCRIPTION : THIS APPLICATION HAS TO BE SENT BY E-MAIL NOT LATER THAN MARCH 31, 1997 TO cargese at cpt.univ-mrs.fr CE FORMULAIRE DOIT ETRE RENVOYE PAR COURRIER ELECTRONIQUE AU PLUS TARD LE 31 MARS 1997 A cargese at cpt.univ-mrs.fr You will receive an answer by e-mail before April 30. Vous recevrez une reponse avant le 30 avril. e-mail contact for special requests only / contact par courrier electronique pour des problemes particuliers: cargese at cpt.univ-mrs.fr From ruderman at salk.edu Tue Feb 18 14:38:21 1997 From: ruderman at salk.edu (Dan Ruderman) Date: Tue, 18 Feb 1997 11:38:21 -0800 Subject: Preprint available Message-ID: <330A052D.745F682F@salk.edu> The following preprint is available: Origins of scaling in natural images Daniel L. Ruderman One of the most robust qualities of our visual world is the scale-invariance of natural images. Not only has scaling been found in different visual environments, but the phenomenon also appears to be calibration independent. This paper proposes a simple property of natural images which explains this robustness: They are collages of regions corresponding to statistically independent ``objects''. Evidence is provided for these objects having a power-law distribution of sizes within images, from which follows scaling in natural images. It is commonly suggested that scaling instead results from edges, each with power spectrum 1/k^2. This hypothesis is refuted by example. (To appear in Vision Research) It may be found at my website: http://quake.usc.edu/~dlr/papers.html -- Dr. Dan Ruderman ruderman at sloan.salk.edu The Salk Institute (VCL) 10010 N. Torrey Pines Road La Jolla, CA 92037 (619) 453-4100 x2025 FAX: (619) 455-7933 From pelillo at dsi.unive.it Wed Feb 19 03:33:05 1997 From: pelillo at dsi.unive.it (Marcello Pelillo) Date: Wed, 19 Feb 1997 09:33:05 +0100 (MET) Subject: EMMCVPR'97 (Venice) - Program Message-ID: <199702190833.JAA16875@oink.dsi.unive.it> A non-text attachment was scrubbed... Name: not available Type: text Size: 8976 bytes Desc: not available Url : https://mailman.srv.cs.cmu.edu/mailman/private/connectionists/attachments/00000000/b1f43ccc/attachment-0001.ksh From gareth at stat.stanford.edu Wed Feb 19 17:25:57 1997 From: gareth at stat.stanford.edu (Gareth James) Date: Wed, 19 Feb 1997 14:25:57 -0800 (PST) Subject: Paper on Generalized Bias/Variance Decomposition Message-ID: The following paper is available in postscript format at http://stat.stanford.edu/~gareth under my papers section. Generalizations of the Bias/Variance Decomposition for Prediction Error Gareth James and Trevor Hastie (Stanford University) The bias and variance of a real valued random variable, using squared error loss, are well understood. However because of recent developments in classification techniques it has become desirable to extend these concepts to general random variables and loss functions. The 0-1 (misclassification) loss function with categorical random variables has been of particular interest. We explore the concepts of variance and bias and develop a decomposition of the prediction error into functions of the systematic and variable parts of our predictor. After providing some examples we conclude with a discussion of the various definitions that have been proposed. From carmesin at schoner.physik.uni-bremen.de Thu Feb 20 11:09:05 1997 From: carmesin at schoner.physik.uni-bremen.de (Hans-Otto Carmesin) Date: Thu, 20 Feb 1997 17:09:05 +0100 Subject: Theory and Experiment papers Message-ID: <199702201609.RAA03423@schoner.physik.uni-bremen.de> Two corresponding experimental and theoretical papers concerning -- Continuous Phase Transitions in Multistable Perception: Neural Network and Experiment -- are now available. Based on a neural network theory, we predicted a continuous phase transition for multistable perception (H.-O. Carmesin: Theorie neuronaler Adaption, Koester, Berlin 1994,1996). Then we discovered this continuous transition experimentally, observed additional dynamical effects at the sensitive continuous transition and improved the neural network model accordingly. The full neural network theory is presented in the paper - A neural network model for stroboscopic alternative motion - by Hans-Otto Carmesin and Stefan Arndt in Biological Cybernetics 75, 239-251 (1996). ABSTRACT: A neural network which models multistable perception is presented. The network consists of sensor and inner neurons. The dynamics is established by a stochastic neuronal dynamics, a formal Hebb-type coupling dynamics and a resource mechanism that corresponds to saturation effects in perception. >From this a system of coupled differential equations is derived and analyzed. Single stimuli are bound to exactly one percept, even in ambiguous situations where multistability occurs. The network exhibits discontinuous as well as continuous phase transitions and models various empirical findings, including the percepts of succession, alternative motion and simultaneity; the percept of oscillation is explained by oscillating percepts at a continuous phase transition. The experimental methods and results are presented in the paper - Continuous phase transitions in the perception of multistable visual patterns - by Peter Kruse, Hans-Otto Carmesin, Lars Pahlke, Daniel Strber and Michael Stadler in Biological Cybernetics 75, 321-330 (1996). ABSTRACT: The phenomenon of alternative stroboscopic motion exhibits five different percepts that are seen with an increase in frequency of presentation: (a) succession, (b) fluttering motion, (c) reversible clockwise and counterclockwise motion, (d) oppositional motion and (e) simultaneity. From a synergetic point of view the increase in frequency is a control parameter and different percepts are order parameters with phase transitions in between. The neural network theory of Carmesin and Arndt is applied to receive predictions about hysteresis and phase transitions between these order parameters. Empirical data show the different motion percepts (b), (c) and (e) have lognormal distributions. Following the theoretical model, it is argued that there are three different phases, (a), (c) and (e), with two continuous phase transitions, (b) and (d), between them. The experimental data substantially match the theoretical assumptions. Some free reprints are available and can be requested via email from carmesin at theo.physik.uni-bremen.de. Moreover, two related papers are available online in the www at http://schoner.physik.uni-bremen.de/~carmesin/docs/zkw9505.ps and http://schoner.physik.uni-bremen.de/~carmesin/docs/cyber2u.ps. Hans-Otto Carmesin Institute for Theoretical Physics, University Bremen, 28334 Bremen, Germany, WWW: http://schoner.physik.uni-bremen.de/~carmesin/ From prefenes at lbs.ac.uk Fri Feb 21 19:55:04 1997 From: prefenes at lbs.ac.uk (Paul Refenes) Date: Fri, 21 Feb 1997 19:55:04 BST Subject: Preprints available - Neural Networks in the Capital Markets Message-ID: <5E6F0DD62D7@neptune.lbs.ac.uk> Neural Networks in the Capital Markets The following NNCM-96 pre-prints are now available on request. Please send your postal address to: boguntula at lbs.lon.ac.uk =================================================== NEURAL MODEL IDENTIFICATION, VARIABLE SELECTION AND MODEL ADEQUACY A-P. N. REFENES, A. D. ZAPRANIS AND J. UTANS Department of Decision Science London Business School Regents Park, London, NW1 4SA, UK In recent years an impressive array of publications have appeared claiming considerable successes of neural networks in modeling financial data but skeptical practitioners and statisticians are still raising the question of whether neural networks really are "a major breakthrough or just a passing fad". A major reason for this is the lack of procedures for performing tests for mispecified models, and tests of statistical significance for the various parameters that have been estimated, which makes it difficult to assess the model's significance and the possibility that any short term successes that are reported might be due to "data mining". In this paper we describe a methodology for neural model identification which facilitates hypothesis testing at two levels: model adequacy and variable significance. The methodology includes a model selection procedure to produce consistent estimators, a variable selection procedure based on variable significance testing and a model adequacy procedure based on residuals analysis. ================================================ SPECIFICATION TESTS FOR NEURAL NETWORKS: A CASE STUDY IN TACTICALASSET ALLOCATION A. D. ZAPRANIS, J. UTANS, A-P. N. REFENES Department of Decision Science London Business School Regents Park, London, NW1 4SA, UK A case study in tactical asset allocation is used to introduce a methodology for neural model identification including model specification and variable selection. Neural models are contrasted to multiple linear regression on the basis of model identification. The results indicate the presence of non-linear relationships between the economic variables and asset class returns. ============================================= From harnad at cogsci.soton.ac.uk Sun Feb 23 12:11:28 1997 From: harnad at cogsci.soton.ac.uk (Stevan Harnad) Date: Sun, 23 Feb 97 17:11:28 GMT Subject: Psycoloquy Call for Papers Message-ID: <18205.9702231711@cogsci.ecs.soton.ac.uk> PSYCOLOQUY CALL FOR PAPERS PSYCOLOQUY is a refereed electronic journal (ISSN 1055-0143) sponsored on an experimental basis by the American Psychological Association and currently estimated to reach a readership of 50,000. PSYCOLOQUY publishes reports of new ideas and findings on which the author wishes to solicit rapid peer feedback, international and interdisciplinary ("Scholarly Skywriting"), in all areas of psychology and its related fields (biobehavioral science, cognitive science, neuroscience, social science, etc.). All contributions are refereed. Target article length should normally not exceed 500 lines [c. 4500 words]. Commentaries and responses should not exceed 200 lines [c. 1800 words]. All target articles, commentaries and responses must have (1) a short abstract (up to 100 words for target articles, shorter for commentaries and responses), (2) an indexable title, (3) the authors' full name(s), institutional address(es) and URL(s). In addition, for target articles only: (4) 6-8 indexable keywords, (5) a separate statement of the authors' rationale for soliciting commentary (e.g., why would commentary be useful and of interest to the field? what kind of commentary do you expect to elicit?) and (6) a list of potential commentators (with their email addresses). All paragraphs should be numbered in articles, commentaries and responses (see format of already published articles in the PSYCOLOQUY archive; line length should be < 80 characters, no hyphenation). It is strongly recommended that all figures be designed so as to be screen-readable ascii. If this is not possible, the provisional solution is the less desirable hybrid one of submitting them as .gif .jpeg .tiff or postscript files (or in some other universally available format) to be printed out locally by readers to supplement the screen-readable text of the article. PSYCOLOQUY also publishes multiple reviews of books in any of the above fields; these should normally be the same length as commentaries, but longer reviews will be considered as well. Book authors should submit a 500-line self-contained Precis of their book, in the format of a target article; if accepted, this will be published in PSYCOLOQUY together with a formal Call for Reviews (of the book, not the Precis). The author's publisher must agree in advance to furnish review copies to the reviewers selected. Authors of accepted manuscripts assign to PSYCOLOQUY the right to publish and distribute their text electronically and to archive and make it permanently retrievable electronically, but they retain the copyright, and after it has appeared in PSYCOLOQUY authors may republish their text in any way they wish -- electronic or print -- as long as they clearly acknowledge PSYCOLOQUY as its original locus of publication. However, except in very special cases, agreed upon in advance, contributions that have already been published or are being considered for publication elsewhere are not eligible to be considered for publication in PSYCOLOQUY, Please submit all material to psyc at pucc.princeton.edu http://www.princeton.edu/~harnad/psyc.html http://cogsci.soton.ac.uk/psyc ftp://ftp.princeton.edu/pub/harnad/Psycoloquy ftp://cogsci.soton.ac.uk/pub/harnad/Psycoloquy gopher://gopher.princeton.edu/11/.libraries/.pujournals news:sci.psychology.journals.psycoloquy ---------------------------------------------------------------------- CRITERIA FOR ACCEPTANCE: To be eligible for publication, a PSYCOLOQUY target article should not only have sufficient conceptual rigor, empirical grounding, and clarity of style, but should also offer a clear rationale for soliciting Commentary. That rationale should be provided in the author's covering letter, together with a list of suggested commentators. A target article can be (i) the report and discussion of empirical research; (ii) an theoretical article that formally models or systematizes a body of research; or (iii) a novel interpretation, synthesis, or critique of existing experimental or theoretical work. Rrticles dealing with social or philosophical aspects of the behavioral and brain sciences are also eligible.. The service of Open Peer Commentary will be primarily devoted to original unpublished manuscripts. However, a recently published book whose contents meet the standards outlined above may also be eligible for Commentary. In such a Multiple Book Review, a comprehensive, 500-line precis by the author is published in advance of the commentaries and the author's response. In rare special cases, Commentary will also be extended to a position paper or an already published article dealing with particularly influential or controversial research. Submission of an article implies that it has not been published or is not being considered for publication elsewhere. Multiple book reviews and previously published articles appear by invitation only. The Associateship and professional readership of PSYCOLOQUY are encouraged to nominate current topics and authors for Commentary. In all the categories described, the decisive consideration for eligibility will be the desirability of Commentary for the submitted material. Controversially simpliciter is not a sufficient criterion for soliciting Commentary: a paper may be controversial simply because it is wrong or weak. Nor is the mere presence of interdisciplinary aspects sufficient: general cybernetic and "organismic" disquisitions are not appropriate for PSYCOLOQUY. Some appropriate rationales for seeking Open Peer Commentary would be that: (1) the material bears in a significant way on some current controversial issues in behavioral and brain sciences; (2) its findings substantively contradict some well-established aspects of current research and theory; (3) it criticizes the findings, practices, or principles of an accepted or influential line of work; (4) it unifies a substantial amount of disparate research; (5) it has important cross-disciplinary ramifications; (6) it introduces an innovative methodology or formalism for consideration by proponents of the established forms; (7) it meaningfully integrates a body of brain and behavioral data; (8) it places a hitherto dissociated area of research into an evolutionary or ecological perspective; etc. In order to assure communication with potential commentators (and readers) from other PSYCOLOQUY specialty areas, all technical terminology must be clearly defined or simplified, and specialized concepts must be fully described. NOTE TO COMMENTATORS: The purpose of the Open Peer Commentary service is to provide a concentrated constructive interaction between author and commentators on a topic judged to be of broad significance to the biobehavioral science community. Commentators should provide substantive criticism, interpretation, and elaboration as well as any pertinent complementary or supplementary material, such as illustrations; all original data will be refereed in order to assure the archival validity of PSYCOLOQUY commentaries. Commentaries and articles should be free of hyperbole and remarks ad hominem. STYLE AND FORMAT FOR ARTICLES AND COMMENTARIES TARGET ARTICLES: should not exceed 500 lines (~4500 words); commentaries should not exceed 200 lines (1800 words), including references. Spelling, capitalization, and punctuation should be consistent within each article and commentary and should follow the style recommended in the latest edition of A Manual of Style, The University of Chicago Press. It may be helpful to examine a recent issue of PSYCOLOQUY. All submissions must include an indexable title, followed by the authors' names in the form preferred for publication, full institutional addresses and electronic mail addresses, a 100-word abstract, and 6-12 keywords. Tables and diagrams should be made screen-readable wherever possible (if unavoidable, printable postscript files may contain the graphics separately). All paragraphs should be numbered, consecutively. No line should exceed 72 characters, and a blank line should separate paragraphs. REFERENCES: Bibliographic citations in the text must include the author's last name and the date of publication and may include page references. Complete bibliographic information for each citation should be included in the list of references. Examples of correct style are: Brown(1973); (Brown 1973); Brown 1973; 1978); (Brown 1973; Jones 1976); (Brown & Jones 1978); (Brown et al. 1978). References should be typed on a separate sheet in alphabetical order in the style of the following examples. Do not abbreviate journal titles. Kupfermann, I. & Weiss, K. (1978) The command neuron concept. Behavioral and Brain Sciences 1:3-39. Dunn, J. (1976) How far do early differences in mother-child relations affect later developments? In: Growing point in ethology, ed. P. P. G. Bateson & R. A. Hinde, Cambridge University Press. Bateson, P. P. G. & Hinde, R. A., eds. (1978) Growing points in ethology, Cambridge University Press. EDITING: PSYCOLOQUY reserves the right to edit and proof all articles and commentaries accepted for publication. Authors of articles will be given the opportunity to review the copy-edited draft. Commentators will be asked to review copy-editing only when changes have been substantial. --------------------------------------------------------------------------- Prof. Stevan Harnad psyc at pucc.princeton.edu Editor, Psycoloquy phone: +44 1703 594-583 fax: +44 1703 593-281 Department of Psychology http://cogsci.soton.ac.uk/psyc University of Southampton http://www.princeton.edu/~harnad/psyc.html Highfield, Southampton ftp://ftp.princeton.edu/pub/harnad/Psycoloquy SO17 1BJ UNITED KINGDOM ftp://cogsci.soton.ac.uk/pub/harnad/Psycoloquy news:sci.psychology.journals.psycoloquy gopher://gopher.princeton.edu/11/.libraries/.pujournals Sponsored by the American Psychological Association (APA) From payman at u.washington.edu Mon Feb 24 20:35:45 1997 From: payman at u.washington.edu (Payman Arabshahi) Date: Mon, 24 Feb 1997 17:35:45 -0800 (PST) Subject: CIFEr'97 Tutorials - New York, March 23, 1997 Message-ID: <199702250135.RAA01799@saul7.u.washington.edu> Computational Intelligence in Financial Engineering Conference CIFEr'97 March 23-25, 1997 Crowne Plaza Manhattan, New York City http://www.ieee.org/nnc/cifer97 Registration information: Barbara Klemm CIFEr'97 Secretariat Meeting Management 2603 Main Street, Suite # 690 Irvine, California 92714 Tel: (714) 752-8205 or (800) 321-6338 Fax: (714) 752-7444 Email: Meetingmgt at aol.com TUTORIALS ---------------------------------------------------------------------------- Risk Management Jan W. Dash, Ph.D. Director Quantitative Analysis Global Risk Management Smith Barney This tutorial will cover 1) characterization of risks in finance: market risk (interest rates, FX rates, equity indices, spreads), trading risk, systems risk (software, hardware, vendors), model risk, and 2) quantitative measurement of risk: the Greeks (Delta, Gamma, Vega), the partial Greeks (Ladders), the new Greeks (Exotics), dollars at risk (n-Sigma analysis), correlations, static scenario analysis, dynamic scenario analysis, Monte Carlo risk analysis, beginnings of risk standards, DPG, Risk Metrics, and 3) case study of risk: the Viacom CVR Options and 4) pricing and hedging for interest rate derivatives. ---------------------------------------------------------------------------- An Introduction to OTC Derivatives and Their Applications John F. Marshall, Ph.D. Executive Director International Association of Financial Engineers This tutorial is for persons with little prior exposure to derivative instruments. It will focus on the basic products, how they trade, and how they are used. It will be largely non-quantitative. The tutorial will examine how derivatives are used by financial engineers for risk management purposes, investment purposes, cash flow management, and creating structured securities. The use of derivatives to circumvent market imperfections, such as asymmetric taxes and transaction costs, will also be demonstrated. The primary emphasis of the tutorial will be swaps (including interest rate swaps, currency swaps, commodity swaps, equity swaps, and macroeconomic swaps). Applications of OTC options, including caps and floors and digital options will also be examined, but to a lesser extent. ---------------------------------------------------------------------------- GARCH Modeling of Financial Time Series R. Douglas Martin, Ph.D. Professor of Statistics, University of Washington Chief Scientist, Data Analysis Products Division of MathSoft, Inc. This tutorial provides an introduction to univariate and multivariate generalized autoregressive heteroscedastic (GARCH) modeling of financial returns time series data, with a focus on modeling conditional volatilities and correlations. Basic aspects of the various models are discussed, including: conditions for stationarity, optimization techniques for maximum likelihood estimation of the models, use of the estimated conditional standard deviations for value-at-risk calculations and options pricing, use of conditional correlations in obtaining conditional volatilities for portfolios. Examples are provided using the S+GARCH object-oriented toolkit for GARCH modeling. ---------------------------------------------------------------------------- Time Series Tools for Finance Andreas Wiegend, Ph.D. Professor, Stern School of Business, New York University This tutorial presents a unifying view of the recent advances of neuro-fuzzy, and other machine learning techniques for time series and finance. It is given jointly by Prof. Andreas Wiegend (Stern School of Business, NYU), and Dr. Georg Zimmerman (Siemens AG, Munich), and presents both conceptual aspects of time series modeling, specific tricks for financial engineering problems, and software engineering aspects for building a trading system. ---------------------------------------------------------------------------- An Introduction to Evolutionary Computation David B. Fogel, PhD Chief Scientist, Natural Selection, Inc., La Jolla Evolutionary computation encompasses a broad field of optimization algorithms that can be applied to diverse, difficult real-world problems. It is particularly useful in addressing stochastic, nonlinear, and time-varying optimization problems, including those arising in financial engineering. This tutorial will provide background on the inspiration, history, and the practical application of evolutionary computation to problems typical of those encountered in financial engineering. ---------------------------------------------------------------------------- Models for Stochastic Volatility: Some Recent Developments Nuno Cato Professor, New Jersey Institute of Technology, Newark Pedro J. F. de Lima Professor, The Johns Hopkins University, Baltimore In this tutorial, we will firstly discuss the importance of modeling stock market's volatility. Secondly, we will review the basic properties of GARCH- type and SV-type models and some of their most successful extensions, namely the SWitching ARCH (SWARCH) models. The performance of these models will be illustrated with some real data examples. Thirdly, we will discuss some problems with the estimation of these models and with their use for risk forecasting. Fourthly, we will describe some recent research and some novel extensions to these models, such as the Long-Memory Stochastic Volatility (LMSV) and the SWitching Stochastic Volatility (SWSV) models. By using examples from recent stock market behavior we illustrate the capabilities and shortcomings of these new modeling and forecasting tools. ---------------------------------------------------------------------------- From tlindroo at bennet.dc.tutech.fi Tue Feb 25 04:31:38 1997 From: tlindroo at bennet.dc.tutech.fi (Tommi Lindroos) Date: Tue, 25 Feb 1997 11:31:38 +0200 Subject: EANN '97: Invitation for participation Message-ID: <199702250931.LAA35528@bennet.dc.tutech.fi> Sorry for this unsolicited mail. It is being sent to you because you are apparently working in or interested in the field of neural networks. Most of these addresses are taken from neural network societies, and participants of earlier EANN conferences. Please let us know if your address should not be in here. International Conference on Engineering Applications of Neural Networks (EANN '97) Royal Institute of Technology Stockholm, Sweden 16-18 June 1997 INVITATION FOR PARTICIPATION AND PROGRAM OUTLINE The International Conference on Engineering Applications of Neural Networks (EANN '97) is the third conference in the series. The two earlier ones were held near Helsinki in 1995 and in London in 1996. The conference is a forum for presenting the latest results on neural network applications in technical fields. Over a hundred papers from 27 countries have been accepted for oral and poster presentations after a review of the abstracts. Some more information on the conference EANN '97 is available on the world wide web site at http://www.abo.fi/~abulsari/EANN97.html, and on last two conferences at http://www.abo.fi/~abulsari/EANN95.html and EANN96.html Contact details E-mail address : eann97 at kth.se Address : EANN '97 SEA PL 953, FIN 20101 Turku 10, Finland The conference has been organised with cooperation from Systems Engineering Association AB Nonlinear Solutions OY and Royal Institute of Technology Conference chairmen: Hans Liljenstrm and Abhay Bulsari Conference venue: Department of Numerical Analysis (NADA), Royal Institute of Technology (KTH), Osquars backe 2, 10044 Stockholm Registration information The conference fee will be SEK 4148 (SEK 3400 excluding VAT) until 28 February, and SEK 4978 (SEK 4080 excluding VAT) after that. The conference fee includes attendance to the conference and the proceedings. If your organisation (university or company or institute) has a VAT registration from a European Union country other than Finland, then your VAT number should be mentioned on the bank transfer as well as the registration form, and VAT need not be added to the conference fee. The correct conference fee amount should be received in the account number 207 799 342, Svenska Handelsbanken International, Stockholm branch. It can be paid by bank transfer, with all expenses paid by the sender, to "EANN Conference". To avoid extra bureaucracy and correction of the amount at the registration desk, make sure that you have taken care of the bank transfer fees. It is essential to mention the name of the participant with the bank transfer. If you need to pay it in another way (bank drafts, Eurocheques, postal order; no credit cards), please contact us at eann97 at kth.se. Invoicing will cost SEK 100. The tentative program outline is as on the following page. The detailed program will be prepared in the end of April. PROGRAM OUTLINE Sunday, 15 June 1600-1800 Registration Room E1 Room E2 Monday, 16 June 0830 Opening 0845 Vision (1) Control Systems (1) 1200 --- lunch break --- 1330 Vision (2) Control Systems (2) 1630 Discussion session Discussion session on Control on Vision Tuesday, 17 June 0830 Process Engineering Biomedical Engineering 1130 --- lunch break --- 1300 Metallurgy Mechanical Engineering 1530 Industrial Panel Discussion Wednesday, 18 June 0830 Hybrid systems Special applications 1130 --- lunch break --- 1300 Electrical/Electronics General Applications 1800 Closing In addition, there will be a poster session on Wednesday, and possibly an evening program on Tuesday. The indicated times are approximate and changes are still possible. Coffee breaks are not indicated here. International Conference on Engineering Applications of Neural Networks (EANN '97) Stockholm, Sweden 16-18 June 1997 Registration form Surname First Name Affiliation (name of the university/company/organisation) E-mail address Postal Address City Country Fax Have you submitted one or more abstracts ? Y/N Abstract number(s) Registration fee sent (amount) SEK ____________ by bank transfer number _______________ from Bank ______________ VAT registration number Date registration fee sent Date registration form sent Any special requirements ? --- END --- From karaali at ukraine.corp.mot.com Tue Feb 25 15:24:44 1997 From: karaali at ukraine.corp.mot.com (Orhan Karaali) Date: Tue, 25 Feb 1997 14:24:44 -0600 Subject: Neural Network Intern For Speech Recognition Message-ID: <199702252024.OAA18339@fiji.mot.com> NEURAL NETWORK INTERN FOR SPEECH RECOGNITION Motorola's Chicago Corporate Research Laboratories is currently seeking a Ph.D. student to join the Speech Synthesis and Machine Learning Group as an intern in its Speech Processing Systems Research Laboratory in Schaumburg, Illinois, for spring 1997. The internship will last at least three months. The Speech Synthesis and Machine Learning Group has developed innovative neural network and signal processing technologies for speech synthesis and speech recognition applications. The intern will work on a component of an HMM/neural network hybrid speech recognizer. The duties of the position include applied research, software development, and conducting experiments with speech data sets. Innovation in research, application of technology and a high level of motivation is the standard for all members of the team. The individual should be in an advanced stage of a Ph.D. program in EE, CS or a related discipline. The ability to work within a group to quickly implement and evaluate algorithms in a rapid research/development cycle is essential. Strong programming skills in the "C" language and solid knowledge of neural networks are required. Background in any of speech processing, statistical techniques, decision trees, and genetic algorithms is highly desirable. Please send text-format resume and cover letter to Orhan Karaali, karaali at mot.com. Motorola is an equal opportunity/affirmative action employer. We welcome and encourage diversity in our workforce. From ejua71 at tattoo.ed.ac.uk Wed Feb 26 12:57:35 1997 From: ejua71 at tattoo.ed.ac.uk (J A Bullinaria) Date: Wed, 26 Feb 97 17:57:35 GMT Subject: Conference announcement: NCPW4 Message-ID: <9702261757.ab24625@uk.ac.ed.tattoo> 4th Neural Computation and Psychology Workshop Connectionist Representations : Theory and Practice University of London, England Wednesday 9th April - Friday 11th April 1997 We have recently solicited, recieved, reviewed and accepted 36 abstracts for presentation as talks at this workshop. We now invite additional participants to register, attend and listen to these talks. AIMS AND OBJECTIVES This workshop is the fourth in a series, following on from the first at the University of Wales, Bangor (with theme "Neurodynamics and Psychology"), the second at the University of Edinburgh, Scotland ("Memory and Language") and the third at the University of Stirling, Scotland ("Perception"). The general aim is to bring together researchers from such diverse disciplines as artificial intelligence, applied mathematics, cognitive science, computer science, neurobiology, philosophy and psychology to discuss their work on the connectionist modelling of psychology. This years workshop is being hosted jointly by members of the Psychology Departments of Birkbeck College London and University College London. As in previous years there will be a theme to the workshop. We think that this years theme is sufficiently wide ranging and important that researchers in all areas of Neural Computation and Psychology will find it relevant and have something to say on the subject. The theme is to be: "Connectionist Representations : Theory and Practice". This covers many important issues ranging from the philosophical (such as the grounding problem) to the physiological (what can connectionist representations tell us about real neural systems) to the technical (such as what is necessary to get specific models to work). As in previous years we aim to keep the workshop fairly small, informal and single track. As always, participants bringing expertise from outside the UK are particularly welcome. SPEAKERS INCLUDE: Roland Baddeley (Oxford) Dennis Norris (APU Cambridge) Tony Browne (Mid Kent) Mike Page (APU Cambridge) Neil Burgess (UCL) Malti Patel (Sydney, Australia) Morten Christiansen (S. Calif.) Tim Shallice (UCL) Robert French (Liege, Belgium) Leslie Smith (Stirling) Peter Hancock (Stirling) John G. Taylor (King's London) Glyn Humphreys (Birmingham) Chris Thornton (Sussex) Geoff Goodhill (Georgetown, DC) Janet Vousden (Warwick) Our web site has a complete listing of all the talks and abstracts. REGISTRATION, FOOD AND ACCOMMODATION The workshop will be held in University College London, which is situated in the centre of London, near the British Museum and within easy walking distance of the West End and many of London's major attractions. The conference registration fee (which includes lunch and morning and afternoon tea/coffee each day) is 60 pounds. A special conference dinner (optional) is planned for the Thursday evening costing 20 pounds. Accommodation can be arranged in student residences or in local hotels, according to budget. The conference/accommodation area is easily accessible by the London Underground system ("The Tube"), with direct lines from London Heathrow Airport and all the major intercity train stations. Full registration and accommodation information is available at the conference web site: "http://prospero.psychol.ucl.ac.uk/ncpw4/". ORGANISING COMMITTEE John Bullinaria (Birkbeck College London) Dave Glasspool (University College London) George Houghton (University College London) CONTACT DETAILS Workshop email address for all correspondence: ncpw4 at psychol.ucl.ac.uk Workshop web page: http://prospero.psychol.ucl.ac.uk/ncpw4/ John Bullinaria, NCPW4, Centre for Speech and Language, Department of Psychology, Birkbeck College, Malet Street, London WC1E 7HX, UK. Phone: +44 171 631 6330, Fax: +44 171 631 6587 Email: j.bullinaria at psyc.bbk.ac.uk Dave Glasspool, NCPW4, Department of Psychology, University College London, Gower Street, London WC1E 6BT, UK. Phone: +44 171 380 7777 Xtn. 5418. Fax: +44 171 436 4276 Email: d.glasspool at psychol.ucl.ac.uk George Houghton, NCPW4, Department of Psychology, University College London, Gower Street, London WC1E 6BT, UK. Phone: +44 171 380 7777 Xtn. 5394. Fax: +44 171 436 4276 Email: g.houghton at psychol.ucl.ac.uk From mel at quake.usc.edu Thu Feb 27 01:20:18 1997 From: mel at quake.usc.edu (Bartlett Mel) Date: Thu, 27 Feb 1997 14:20:18 +0800 Subject: UCSD/USC Joint Symposium on Neural Computation Message-ID: <9702272220.AA00232@quake.usc.edu> CALL FOR PRESENTATIONS --- 4th Annual Joint Symposium on Neural Computation --- Co-sponsored by Institute for Neural Computation University of California, San Diego and Biomedical Engineering Department and Neuroscience Program University of Southern California to be hosted at The University of Southern California University Park Campus Saturday, May 17, 1997 9:00 a.m. to 6:00 p.m. In 1994, the Institute for Neural Computation at UCSD hosted the first Joint Symposium on Neural Computation with Caltech that brought together students and faculty for a day of short presentations. This year USC will be the site for the fourth Symposium and will feature as keynote speaker: Prof. Irving Biederman Departments of Psychology and Computer Science and the Neuroscience Program University of Southern California "Shape Representation in Mind and Brain" Submissions will be open to members of the Computational Neuroscience community of Southern California. Given the larger constituency than in previous years, authors are invited to contribute 300 word abstracts, which will be reviewed by a program committee consisting of the USC organizers and representatives of the INC. The contributed program will consist of 15 minute oral presentations, and posters. Abstracts selected for presentation will be included in the final program. DEADLINE FOR RECEIPT OF ABSTRACTS: March 28, 1997 Submissions should be e-mailed or mailed to Linda Yokote using the form below. Notification of acceptance as oral or poster presentation will be e-mailed to authors by April 18, 1997. A proceedings of short papers will be published by the INC. Contributions to the proceedings, based on both oral and poster presentations, must be submitted no later than May 30, 1997 for timely publication to: Institute for Neural Computation, University of California San Diego, 9500 Gilman Drive DEPT 0523, La Jolla, California 92093-0523. As in previous years, authors will retain copyright to their papers, so that they may be resubmitted elsewhere. Registration and attendance at the Symposium is open to the public. USC Organizing Committee: Dr. Bartlett Mel - Biomedical Engineering Department mel at quake.usc.edu, http://quake.usc.edu/lnc.html Dr. Michael Arbib - Professor of Computer Science and Neurobiology Director of the USC Brain Project arbib at pollux.usc.edu, http:/www-hbp.usc.edu/HBP ---------------------------------------------------------------------------- ---------------------------------------------------------------------------- 1997 JSNC Submission Form - Return to: Linda Yokote yokote at bmsrs.usc.edu (e-mail submissions preferred) US mail: Joint Symposium Biomedical Engineering Department USC, MC 1451 Los Angeles, CA 90089 (213)740-0840, (213)740-0343 fax ---------------------------------------------------------------------------- _____ I would like to attend the Symposium. Registration fee of $25 includes lunch and Proceedings. Checks are payable to the Department of Biomedical Engineering, USC. _____ I would like to give a presentation Title: _______________________________________________________________________ _______________________________________________________________________ Abstract: (300 word abstract goes here) _______________________________________________________________________ Speaker's Name: ______________________________________________________________ Affiliation/Department: ______________________________________________________ Address: _____________________________________________________________________ _____________________________________________________________________ _____________________________________________________________________ Telephone: ________________________ E-mail Address: _____________________ Others who should be listed as co-authors: ______________________________ ______________________________________________________________________________ ______________________________________________________________________________ ______________________________________________________________________________ --- DEADLINE FOR RECEIPT OF ABSTRACTS: March 28, 1997 --- From bishopc at helios.aston.ac.uk Thu Feb 27 05:09:52 1997 From: bishopc at helios.aston.ac.uk (Prof. Chris Bishop) Date: Thu, 27 Feb 1997 10:09:52 +0000 Subject: Isaac Newton Institute Message-ID: <3914.199702271009@sun.aston.ac.uk> Isaac Newton Institute NEURAL NETWORKS AND MACHINE LEARNING A six month programme at the Isaac Newton Institute for Mathematical Sciences, Cambridge, U.K. July to December 1997 Organisers: C M Bishop (Aston), D Haussler (UCSC), G E Hinton (Toronto), M Niranjan (Cambridge), L G Valiant (Harvard) The Isaac Newton Institute for Mathematical Sciences is an international research centre, sponsoring visitor programmes in topics across the whole spectrum of mathematical research. This programme will be an important international event in the field of neural computing. At any one time there will be around 20 to 25 long-term participants, as well as larger numbers of short-term visitors. Seminars are open to all, although long-term participation will be by invitation. Three major conferences are planned to take place during the programme: 1) "Generalization in Neural Networks and Machine Learning" (4 to 15 August). This will be a NATO Advanced Study Institute. 2) "Probabilistic Graphical Models" (1 to 5 September). 3) "Bayesian Methods" (15 to 19 December). In addition, there will be a number of workshops, generally of one week duration. Provisional themes for these include the following: "Learning in computer vision" (6 to 10 October), "HMM hybrid models and protein/DNA modelling" (20 to 24 October), "Applications" (3 to 7 November), "Non-stationarity" (10 to 14 November), "Information geometry" (8 to 12 December). Information about the programme, and a list of participants, can be found at the programme web site: http://www.newton.cam.ac.uk/programs/nnm.html General information about the Newton Institute can be found at: http://www.newton.cam.ac.uk/ To be kept informed of future developments, you can subscribe to the programme information mailing list by sending an e-mail to majordomo at newton.cam.ac.uk with a message whose body contains the line subscribe nnm-list For programme-specific enquires please contact: Christopher M. Bishop Neural Computing Research Group Dept. of Computer Science and Applied Mathematics Aston University, Birmingham B4 7ET, U.K. Tel. +44/0 121 333 4631 Fax. +44/0 121 333 4586 C.M.Bishop at aston.ac.uk From movellan at ergo.ucsd.edu Thu Feb 27 19:31:44 1997 From: movellan at ergo.ucsd.edu (Javier R. Movellan) Date: Thu, 27 Feb 1997 16:31:44 -0800 Subject: TR Announcement Message-ID: <199702280031.QAA29932@ergo.ucsd.edu> The following technical report is available online at http://cogsci.ucsd.edu (follow links to Tech Reports & Software ) Physical copies are also available (see the site for information). From Contour Completion to Image Schemas: A Modern Perspective on Gestalt Psychology. Adrian Robert (Communicated by Martin I. Sereno) Department of Cognitive Science University of California San Diego The Gestalt approach to psychology represents an early but comprehensive and systematic attempt to relate psychological and neural functioning. When the approach was first formulated and actively researched, however, too little was known about brain function to forge a precise and direct connection. As a result, the approach never fulfilled its initial promise of a rigorously founded psychology grounded in physical science and has fallen out of the favor and attention of most contemporary students of the mind. In this paper we re-examine Gestalt psychology with reference to what is currently known of dynamic mechanisms of brain function, particularly by exploring plausible neural substrates of perceptual grouping. We suggest, based on this examination, that although many of the details of the Gestalt proposals are in need of revision, the approach remains fundamentally viable, and the elegant character of its grounding and systematicity make it a valuable framework for organizing present knowledge at both neural and functional levels. From gerda at ai.univie.ac.at Fri Feb 28 04:00:58 1997 From: gerda at ai.univie.ac.at (Gerda Helscher) Date: Fri, 28 Feb 1997 10:00:58 +0100 Subject: New Book: NN and a new AI Message-ID: <33169ECA.2781E494@ai.univie.ac.at> !!! New Book Announcement !!! ================================================= Neural Networks and a New Artificial Intelligence ================================================= edited by Georg Dorffner International Thomson Computer Press, London, 1997 ISBN 1-85032-172-8 About the book: =============== Since the re-birth of interest in artificial neural networks in the mid 1980s, they have become a much-discussed topic, particularly in terms of their real contribution to the explanation and modelling of cognition, as part of the field of artificial intelligence (AI). This edited collection brings together a selection of papers from experts in their field, outlining the concrete contribution that neural computing has made to AI. "Neural Networks and a New Artificial Intelligence" is a collection of arguments, examples and critical elaborations from different views on how and whether neural networks can not only contribute to a better artificial intelligence, but can also revolutionise it by forming the basis for a truly alternative paradigm. Contents: ========= Introduction Part I: General topics - new AI as a whole New AI: naturalness revealed in the study of artificial intelligence (by Erich Prem, Vienna) Representational eclecticism - a foundation stone for the new AI? (by Chris Thornton, Brighton) Part II: concrete approaches and research strategies Towards a connectionist model of action sequences, active vision and breakdowns (by Hugues Bersini, Brussels) Complete autonomous systems: a research strategy for cognitive science (by Rolf Pfeifer and Paul Verschure, Zurich) Radical connectionism - a neural bottom-up approach to AI (by Georg Dorffner, Vienna) Connectionist explanation: taking position in the mind-brain dilemma (by Paul Verschure, Zurich) On growing intelligence (by Jari Vaario, Kyoto, and Setsuo Ohsuga, Tokyo) Part III: Issues in modelling high-level cognition Systematicity and generalization in compositional connectionist representations (by Lars Niklasson, Sk"ovde, and Noel Sharkey, Sheffield) Constructive learning in connectionist semantic networks (by Joachim Diederich and James Hogan, Brisbane) A connectionist Model for the interpretation of metaphors (by Stefan Wermter, Hamburg, and Ruth Hannuschka, Dortmund) Some issues in neural cognitive modelling (by Max Garzon, Memphis) Neural networks and a new AI - questions and answers (with contributions from all the authors) Index