From m.niranjan at dcs.shef.ac.uk Thu Mar 1 11:51:56 2001 From: m.niranjan at dcs.shef.ac.uk (Mahesan Niranjan) Date: Thu, 1 Mar 2001 16:51:56 +0000 (GMT) Subject: Workshop on Geometric Computations In-Reply-To: <200103011100.LAA13013@padley.dcs.shef.ac.uk> Message-ID: +++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++ UNCERTAINTY IN GEOMETRIC COMPUTATIONS, 5-6 July 2001, Sheffield, England Invited Speakers (will be expanded) : Shun-ichi Amari (RIKEN, Japan), Andrew Blake (Microsoft, UK), Adrian Bowyer (Bath, UK), Alan Edelman (MIT, USA), Robin Forrest (East Anglia, UK), Nicholas Higham (Manchester, UK), Dinesh Manocha (North Carolina, USA), Si Wu (Sheffield, UK) Organisers: Joab Winkler and Mahesan Niranjan Department of Computer Science The University of Sheffield, UK. The representation and management of uncertainty is an important issue in several different disciplines, such as numerical problems in computer graphics that occur when calculating the intersection curve of two surfaces, high performance pattern classification in a feature space, and the study of families of probability distributions in information geometry. The aim of this two-day workshop is to explore the underlying geometric theme that is common to these diverse disciplines. The workshop will consist of a number of invited contributions of a tutorial nature covering the different topics, contributed papers from participants and discussion sessions that explore the connections. Contributions will be published by Kluwer in an edited volume. The workshop is sponsored by the EPSRC and LMS, and financial support is available to cover costs of UK based graduate students. The total number of participants is limited to 70. One page abstracts are invited from potential participants. Please submit electronically (postscript, PDF or plain text) to Dr Joab Winkler Deadline for Abstracts: 15 April 2001 For further information see: http://www.shef.ac.uk/~geom2001/ +++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++ From wolfskil at MIT.EDU Thu Mar 1 15:51:02 2001 From: wolfskil at MIT.EDU (Jud Wolfskill) Date: Thu, 1 Mar 2001 15:51:02 -0500 Subject: book announcement--Spirtes Message-ID: I thought readers of the Connectionist List might be interested in this book. For more information please visit http://mitpress.mit.edu/promotions/books/SPICHF00. Thank you! Best, Jud Causation, Prediction, and Search second edition Peter Spirtes, Clark Glymour, and Richard Scheines What assumptions and methods allow us to turn observations into causal knowledge, and how can even incomplete causal knowledge be used in planning and prediction to influence and control our environment? In this book Peter Spirtes, Clark Glymour, and Richard Scheines address these questions using the formalism of Bayes networks, with results that have been applied in diverse areas of research in the social, behavioral, and physical sciences. The authors show that although experimental and observational study designs may not always permit the same inferences, they are subject to uniform principles. They axiomatize the connection between causal structure and probabilistic independence, explore several varieties of causal indistinguishability, formulate a theory of manipulation, and develop asymptotically reliable procedures for searching over equivalence classes of causal models, including models of categorical data and structural equation models with and without latent variables. The authors show that the relationship between causality and probability can also help to clarify such diverse topics in statistics as the comparative power of experimentation versus observation, Simpson's paradox, errors in regression models, retrospective versus prospective sampling, and variable selection. The second edition contains a new introduction and an extensive survey of advances and applications that have appeared since the first edition was published in 1993. Peter Spirtes is Professor of Philosophy at the Center for Automated Learning and Discovery, Carnegie Mellon University. Clark Glymour is Alumni University Professor of Philosophy at Carnegie Mellon University and Valtz Family Professor of Philosophy at the University of California, San Diego. He is also Distinguished External Member of the Center for Human and Machine Cognition at the University of West Florida, and Adjunct Professor of Philosophy of History and Philosophy of Science at the University of Pittsburgh. Richard Scheines is Associate Professor of Philosophy at the Center for Automated Learning and Discovery, and at the Human Computer Interaction Institute, Carnegie Mellon University. 7 x 9, 496 pp., 225 illus., cloth ISBN 0-262-19440-6 Adaptive Computation and Machine Learning series A Bradford Book -------------------------------------------------------------------------------- Jud Wolfskill 617.253.2079 phone Associate Publicist 617.253.1709 fax MIT Press wolfskil at mit.edu 5 Cambridge Center http://mitpress.mit.edu Fourth Floor Cambridge, MA 02142 From harnad at coglit.ecs.soton.ac.uk Fri Mar 2 11:15:43 2001 From: harnad at coglit.ecs.soton.ac.uk (Stevan Harnad) Date: Fri, 2 Mar 2001 16:15:43 +0000 (GMT) Subject: Review of: Damasio, Edelman, McGinn, Tomasello and Fodor Message-ID: Review of: Antonio Damasio: The Feeling of What Happens Gerald M. Edelman and Giulio Tononi: A Universe of Consciousness Colin McGinn The Mysterious Flame Michael Tomasello Cultural Origins of Human Cognition Jerry A. Fodor The Mind Doesn't Work That Way To appear in The Sciences New York Academy of Sciences, April 2001 k http://www.cogsci.soton.ac.uk/~harnad/Tp/bookrev.htm -------------------------------------------------------------------- Stevan Harnad harnad at cogsci.soton.ac.uk Professor of Cognitive Science harnad at princeton.edu Department of Electronics and phone: +44 23-80 592-582 Computer Science fax: +44 23-80 592-865 University of Southampton http://www.cogsci.soton.ac.uk/~harnad/ Highfield, Southampton http://www.princeton.edu/~harnad/ SO17 1BJ UNITED KINGDOM From ted.carnevale at yale.edu Sun Mar 4 19:27:01 2001 From: ted.carnevale at yale.edu (Ted Carnevale) Date: Sun, 04 Mar 2001 19:27:01 -0500 Subject: NEURON 2001 Summer Course Message-ID: <3AA2DD55.8343479E@yale.edu> COURSE ANNOUNCEMENT What: "The NEURON Simulation Environment" (NEURON 2001 Summer Course) When: Saturday, June 23, through Wednesday, June 27, 2001 Where: San Diego Supercomputer Center University of California at San Diego, CA Organizers: N.T. Carnevale and M.L. Hines Faculty includes: N.T. Carnevale, M.L. Hines, W.W. Lytton, and T.J. Sejnowski Description: This intensive hands-on course covers the design, construction, and use of models in the NEURON simulation environment. It is intended primarily for those who are concerned with models of biological neurons and neural networks that are closely linked to empirical observations, e.g. experimentalists who wish to incorporate modeling in their research plans, and theoreticians who are interested in the principles of biological computation. The course is designed to be useful and informative for registrants at all levels of experience, from those who are just beginning to those who are already quite familiar with NEURON or other simulation tools. Registration is limited to 20, and the deadline for receipt of applications is Friday, May 25, 2001. For more information see http://www.neuron.yale.edu/sdsc2001/sdsc2001.htm or contact Ted Carnevale Psychology Dept. Box 208205 Yale University New Haven, CT 06520-8205 USA phone 203-432-7363 fax 203-432-7172 email ted.carnevale at yale.edu Supported in part by: National Science Foundation National Institutes of Health National Partnership for Advanced Computational Infrastructure and the San Diego Supercomputer Center Contractual terms require inclusion of the following statement: This course is not sponsored by the University of California. --Ted From juergen at idsia.ch Mon Mar 5 12:24:11 2001 From: juergen at idsia.ch (juergen@idsia.ch) Date: Mon, 5 Mar 2001 18:24:11 +0100 Subject: optimal search algorithm Message-ID: <200103051724.SAA27997@ruebe.idsia.ch> Dear connectionists, Recently Marcus Hutter has developed a very general, asymptotically optimal search method that should be of interest to researchers in the areas of reinforcement learning & neural networks & AI. The method is not limited to machine learning problems though - I believe it will find its way into general computer science textbooks. With the benefit of hindsight I find it amazing that it has remained undiscovered up until the turn of the century. ------------------------------------------------------------------------- The fastest and shortest algorithm for all well-defined problems Marcus Hutter, IDSIA An algorithm M is described that solves any well-defined problem p as quickly as the fastest algorithm computing a solution to p, save for a factor of 5 and low-order additive terms. M optimally distributes resources between the execution of provably correct p-solving programs and an enumeration of all proofs, including relevant proofs of program correctness and of time bounds on program runtimes. M avoids Blum's speed-up theorem by ignoring programs without correctness proof. M has broader applicability and can be faster than Levin's universal search, the fastest method for inverting functions save for a large multiplicative constant. An extension of Kolmogorov complexity and two novel natural measures of function complexity are used to show that the most efficient program computing some function f is also among the shortest programs provably computing f. ftp://ftp.idsia.ch/pub/techrep/IDSIA-16-00.ps.gz ------------------------------------------------------------------------- Juergen Schmidhuber www.idsia.ch From t.c.pearce at leicester.ac.uk Mon Mar 5 13:20:56 2001 From: t.c.pearce at leicester.ac.uk (Tim Pearce) Date: Mon, 5 Mar 2001 18:20:56 -0000 Subject: Postdoctoral Research Positions Message-ID: Three Postdoctoral Research Positions Silicon Olfactory System Implementation We seek applications for three postdoctoral RAs to form the core team of a large EPSRC funded project to develop the worlds first silicon implementation of the olfactory system. Work will be aimed towards producing a micropower, fully integrated, electronic nose, with the capability of adapting to changes in its operating conditions. This is a major project conducted between three UK Universities and, Osmetech PLC a leading international chemical sensing instrumentation entity. There will be various opportunities for presenting work at international conferences, as well as opportunities for travel to our Swiss collaborators at ETH, Zrich. More details on the project are available at http://www.le.ac.uk/eg/tcp1/avlsi/ 1. Computational Neuroscientist will be responsible for developing a detailed computational model of the olfactory bulb, constrained by biological data. The model will be driven by novel ChemFET array technology developed at Warwick and finally will be implemented in aVLSI at Edinburgh. The applicant will ideally have modeling experience in neuronal systems (but not necessarily the olfactory system), a PhD in a related area and strong mathematical skills. For more information on the overall project and this position in particular contact Dr. Tim Pearce, Department of Engineering, University of Leicester, University Road, Leicester LE1 7RH, UK, +44 (0) 116 223 1290, t.c.pearce at le.ac.uk 2. Silicon Sensor/MEMS Specialist will be responsible for the design and fabrication of a novel odour sensing structure based upon microfluidic channels carrying the odour to an array of polymer-coated MOSFET devices. The applicant will have a PhD in electronics and ideally some experience in using device layout and simulation software, such as Tanner Tools or Cadence, Medici, ISE. For more information on the sensors position contact Prof. Julian Gardner, School of Engineering, University of Warwick, Coventry CV4 7AL, UK +44 (0) 24 76523 695, j.w.gardner at warwick.ac.uk 3. Analogue VLSI Engineer. S/He will be responsible for designing the on chip analogue interface to the ChemFET sensor array technology developed at Warwick and implementing the computational model of the olfactory bulb developed at Leicester. The applicant will ideally have some experience in analogue circuit design, good written and verbal communication skills and a PhD in a related subject area. For more details contact Dr. Alister Hamilton, Department of Electronics and Electrical Engineering, University of Edinburgh, Mayfield Road, Edinburgh EH9 3JL, UK, +44 (0) 131 650 5597, Alister.Hamilton at ee.ed.ac.uk Salary range will be on the RA1A scale (16,775 - 25,213 depending upon age and experience). The RAs will be employed for up to three years. Applications should be addressed to the respective academic, including a CV, list of publications and the name of two referees, who should be asked to send their reference to the same address). Applications may also be e-mailed. Closing date: 1st July, 2001. -- T.C. Pearce, PhD URL: http://www.leicester.ac.uk/eg/tcp1/ Lecturer in Bioengineering E-mail: t.c.pearce at leicester.ac.uk Department of Engineering Tel: +44 (0)116 223 1290 University of Leicester Fax: +44 (0)116 252 2619 Leicester LE1 7RH Bioengineering, Transducers and United Kingdom Signal Processing Group From ingber at ingber.com Mon Mar 5 20:30:55 2001 From: ingber at ingber.com (Lester Ingber) Date: Mon, 5 Mar 2001 19:30:55 -0600 Subject: Paper: Probability tree algorithm for general diffusion processes Message-ID: <20010305193055.A17296@ingber.com> The following preprint is available: %A L. Ingber %A C. Chen %A R.P. Mondescu %A D. Muzzall %A M. Renedo %T Probability tree algorithm for general diffusion processes %D 2001 %O URL http://www.ingber.com/path01_pathtree.ps.gz ABSTRACT Motivated by path-integral numerical solutions of diffusion processes, PATHINT, we present a new tree algorithm, PATHTREE, which permits extremely fast accurate computation of probability distributions of a large class of general nonlinear diffusion processes. -- Prof. Lester Ingber http://www.ingber.com/ PO Box 06440 Sears Tower Chicago IL 60606-0440 http://www.alumni.caltech.edu/~ingber/ From terry at salk.edu Wed Mar 7 17:48:31 2001 From: terry at salk.edu (Terry Sejnowski) Date: Wed, 7 Mar 2001 14:48:31 -0800 (PST) Subject: NEURAL COMPUTATION 13:4 Message-ID: <200103072248.f27MmVH58010@kepler.salk.edu> Neural Computation - Contents - Volume 13, Number 4 - April 1, 2001 ARTICLE Sampling Properties of the Spectrum and Coherency of Sequences of Action Potentials M. R. Jarvis and P. P. Mitra NOTE A Novel Spike Distance M. C. W. van Rossum LETTERS On Synchrony of Weakly Coupled Neurons at Low Firing Rates L. Neltner and David Hansel Population Coding with Correlation and an Unfaithful Model Si Wu, Hiroyuki Nakahara and Shun-ichi Amari Metabolically Efficient Information Processing Vijay Balasubramanian, Don Kimber and Michael J. Berry II Effective Neuronal Learning with Ineffective Hebbian Learning Rules Gal Chechik, Isaac Meilijson, and Eytan Ruppin Internal Model Reproduces Anticipatory Neural Activity Roland E. Suri and Wolfram Schultz Blind Source Separation by Sparse Decomposition Michael Zibulevsky and Barak A. Pearlmutter Complexity Pursuit: Separating Interesting Components from Time-Series Aapo Hyvarinen Algebraic Analysis for Non-Identifiable Learning Machines Sumio Watanabe A New On-Line Learning Model Shahar Mendelson ----- ON-LINE - http://neco.mitpress.org/ SUBSCRIPTIONS - 2001 - VOLUME 13 - 12 ISSUES USA Canada* Other Countries Student/Retired $60 $64.20 $108 Individual $88 $94.16 $136 Institution $460 $492.20 $508 * includes 7% GST MIT Press Journals, 5 Cambridge Center, Cambridge, MA 02142-9902. Tel: (617) 253-2889 FAX: (617) 577-1545 journals-orders at mit.edu ----- From doya at isd.atr.co.jp Thu Mar 8 04:21:24 2001 From: doya at isd.atr.co.jp (Kenji Doya) Date: Thu, 8 Mar 2001 18:21:24 +0900 Subject: CREST Workshop on Metalearning and Neuromodulation Message-ID: Dear connectionists, We are organizing the following workshop. The program and the abstracts are posted on the web page. If you are interested, please register as soon as possible. We still have a few openings for poster presentations. Best wishes, Kenji Doya ************************************************************************ CREST WORKSHOP ON METALEARNING AND NEUROMODULATION April 6th and 7th, 2001 Keihanna Plaza, Seika, Kyoto, Japan Sponsored by Metalearning, Neuromodulation, and Emotion Project CREST, Japan Science and Technology Corporation The goal of this workshop is to bring together neuroscientists as well as theorists who work on the regulatory mechanisms of adaptive systems, either biological or artificial. Neuromodulators such as dopamine, serotonin, norepinephrine, and acetylcholine have widespread influences on information processing in the brain. Recent neurobiological studies revealed their specific roles such as prediction of future reward and punishment, regulation of accuracy and diversity of behaviors, and the rate of memory update. Computational studies of 'metalearning,' the way of adapting global parameters and structures in a learning system, can be helpful in understanding the roles and the dynamics of neuromodulators. We invite researchers who are working on the the biological mechanisms of neuromodulators and the computational theory of metalearning, as well as those who are interested in the neurochemical basis and computational mechanisms of the mind. Please visit our home page for details: http://www.isd.atr.co.jp/nip/crest/workshop/ This workshop is scheduled after the 9th International Catecholamine Symposium (http://mc-net.jtbcom.co.jp/ics2001/) held in downtown Kyoto from March 31st to April 5th, 2001. INVITED SPEAKERS Minoru Asada, Osaka University Gary Aston-Jones, University of Pennsylvania Kenji Doya, CREST, JST & ATR International Barry J. Everitt, University of Cambridge Michael Hasselmo, Boston University Okihide Hikosaka, Juntendo University Tadashi Isa, National Institute for Physiological Sciences Shin Ishii, Nara Institute of Science and Technology Sham Kakade, University College London Takashi Matsumoto, Waseda University Toshiyuki Sawaguchi, Hokkaido University Wolfram Schultz, University of Fribourg Yuko Sekino, Gumma University Shigeto Yamawaki, Hiroshima University CALL FOR POSTERS A poster session will be held in the evening of April 6th. Please send 1) title, 2) author(s), 3) affiliation(s), 4) abstract up to 500 words, 5) postal address, 6) phone, 7) fax, and 8) e-mail address of the presenting author by e-mail to the secretariat. REGISTRATION Please send 1) name, 2) affiliation, 3) postal address, 4) phone, 5) fax, and 6) e-mail address by e-mail to the secretariat by March 15, 2001. Registration is free. However, the numbers of attendees will be limited by capacity of the conference room. So please register early. HOTEL RESERVATION Please visit our home page (http://www.isd.atr.co.jp/nip/crest/workshop/) to download a reservation form, and send it by fax directly to a hotel. Note that early April is the high season for cherry blossoms in Kyoto, so please reserve early. SECRETARIAT Naomi Katayama Metalearning, Neuromodulation and Emotion, CREST, JST 2-2-2 Hikaridai, Seika, Soraku, Kyoto 619-0288, Japan Phone: +81-774-95-1251 Fax: +81-774-95-1259 E-mail: nip-info at isd.atr.co.jp URL: http://www.isd.atr.co.jp/nip/crest/ ---- Kenji Doya Information Sciences Division, ATR International; CREST, JST 2-2-2 Hikaridai, Seika, Soraku, Kyoto 619-0288, Japan Phone:+81-774-95-1251; Fax:+81-774-95-1259; http://www.isd.atr.co.jp/~doya From rsun at pc113.cecs.missouri.edu Fri Mar 9 19:09:36 2001 From: rsun at pc113.cecs.missouri.edu (Dr. Ron Sun) Date: Fri, 9 Mar 2001 18:09:36 -0600 Subject: new papers on cognitive models available Message-ID: <200103100009.f2A09aG12197@pc113.cecs.missouri.edu> Announcing several papers on cognitive modeling and cognitive architectures based on hybrid reinforcement learning --- the CLARION model: A paper on cognitive modeling using CLARION: -------------------------------------------------- From sgg at copper.dcs.qmw.ac.uk Mon Mar 12 07:35:57 2001 From: sgg at copper.dcs.qmw.ac.uk (sgg@copper.dcs.qmw.ac.uk) Date: Mon, 12 Mar 2001 12:35:57 +0000 (GMT) Subject: Post-doc positions in computer vision Message-ID: <200103121235.MAA26547@dcs.qmw.ac.uk> COMPUTER VISION RESEARCH: UNDERSTANDING VISUAL BEHAVIOUR DEPARTMENT OF COMPUTER SCIENCE QUEEN MARY, UNIVERSITY OF LONDON U.K. Postdoctoral Research Assistants (2 posts) Applications are invited for two Post-Doctoral Research Assistants to work on a computer vision project funded by EPSRC, DTI and industry on the recognition of visual behaviour in video data. In addition to Queen Mary, University of London, the project consortium consists of a number of industrial partners including Safehouse Technologies, BAA (former British Airport Authorities), BT Labs, and Heritage Protection. You will undertake novel research in dynamic scene understanding, zero-motion recognition, visual events modelling, behaviour profiling and abnormal behaviour recognition. You should have experience in computer vision and statistical learning research. Any experience in processing image sequence data from live video cameras would be an advantage. You should also be competent in programming with C or C++ within X and NT environments. More details about computer vision research at Queen Mary can be found at http://www.dcs.qmw.ac.uk/research/vision/. Both posts are for 2 years and are to start as soon as possible. Salary are in range of 20,865 pounds - 22,599 pounds per annum inclusive, depending on experience. For further details and an application form, please phone the Department of Computer Science, Queen Mary, University of London on 020-7882 5227 quoting reference number 01046, or contact Shaogang Gong at sgg at dcs.qmw.ac.uk or Dennis Parkinson at dennisp at dcs.qmw.ac.uk. Completed application forms should be returned by Wednesday 4th April 2001 to Ms. Gill Carter, Department of Computer Science, Queen Mary, University of London, London, E1 4NS. QUEEN MARY: WORKING TOWARDS EQUAL OPPORTUNITIES From aco-list at iridia0.ulb.ac.be Mon Mar 12 11:34:24 2001 From: aco-list at iridia0.ulb.ac.be (aco-list@iridia0.ulb.ac.be) Date: Mon, 12 Mar 2001 17:34:24 +0100 (CET) Subject: Ant Colony Optimization mailing list Message-ID: <200103121634.RAA21569@iridia0.ulb.ac.be> Dear Colleagues, we would like to announce that a moderated digest about Ant Colony Optimization and Ant Algorithms has been set up. It collects news, information, new ideas and comments about the above-mentioned research area. Main topics: - Ant colony optimization - Ant algorithms - Swarm intelligence and insect behavior - Bio-inspired multi-agent systems Relevant information for the ACO-list are: - Conferences and workshops - Recent pubblications - New products (not just for commercial purposes, but with scientific interest) - Books and articles reviews - Open positions and any other information which could be useful for researchers in this area. For SUBSCRIPTIONS write to: ACO-list-request at iridia.ulb.ac.be For SUBMISSIONS write to: ACO-list at iridia.ulb.ac.be Marco Dorigo and Andrea Roli (list moderators) From Martin.Appl at mchp.siemens.de Mon Mar 12 06:46:12 2001 From: Martin.Appl at mchp.siemens.de (Martin Appl) Date: Mon, 12 Mar 2001 12:46:12 +0100 Subject: Ph.D. thesis available: Model-Based Reinforcement Learning in Continuous Environments Message-ID: <000101c0aaea$0a8dd3b0$0dba178b@mhpaaxsc.mchp.siemens.de> Dear Connectionists, my Ph.D. thesis **************************************************************** Model-Based Reinforcement Learning in Continuous Environments Martin Appl December 2000, Technical University of Munich **************************************************************** is now available at www.martinappl.de . Best regards, Martin Appl ++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++ ABSTRACT Reinforcement learning enables machines to learn from experiences. For example, controllers can learn optimal control strategies trying out different strategies and evaluating the resulting performance of the processes under control. At present reinforcement learning is rarely used for the optimization of complex industrial processes, since the computational requirements of reinforcement learning approaches grow fast as the number of input variables increases. Hence, the rough goal of this thesis is to develop efficient approaches enhancing the range of application of reinforcement learning. The focus of the thesis is on time-discrete control of processes with continuous controlled inputs and continuous measured outputs. A central result of this thesis is a fuzzy model-based reinforcement learning approach. Using this approach control strategies for continuous processes can be efficiently trained. The output of the approach is a Takagi-Sugeno fuzzy system representing the optimal control strategy. A further result of this thesis is a fuzzy model-based exploration strategy. During learning this strategy controls processes in such a way that maximum information is gained. Hence, the number of control cycles required to learn optimal control strategies is significantly reduced. For many control problems it is known a priori which measured quantities are correlated and which are statistically independent. Taking this kind of a priori knowledge into account both the model-based learning approach and the exploration strategy can be significantly sped up, as is also shown in this thesis. A general problem in fuzzy model-based learning is the generation of suitable fuzzy partitions. Defining partitions by hand is not trivial, since fine partitions lead to a large number of states, whereas coarse partitions can be unsuitable for the representation of the optimal control strategy. Therefore, further extensions of the fuzzy model-based learning approach and the model-based exploration strategy are presented in this thesis. The basic idea behind these extensions is to represent internal models by clustered transitions. Based on this compact representation the extended algorithms can automatically determine suitable partitions of the state space. The methods presented in this thesis are applied to tasks from traffic signal control. One task is to select framework signal plans in dependence of traffic conditions. It turned out that the fuzzy model-based approaches outperform existing crisp methods. Furthermore, these methods allow to solve the task in reasonable time. From mieko at isd.atr.co.jp Tue Mar 13 05:06:59 2001 From: mieko at isd.atr.co.jp (Mieko Namba) Date: Tue, 13 Mar 2001 19:06:59 +0900 Subject: Neural Networks 14(3) Message-ID: NEURAL NETWORKS 14(3) Contents - Volume 14, Number 3 - 2001 ------------------------------------------------------------------ LETTERS: New theorems on global convergence of some dynamical systems. T. Chen, S. Amari INVITED ARTICLE: Bayesian approach for neural networks - review and case studies. J. Lampinen, A. Vehtari CONTRIBUTED ARTICLES: ***** Neuroscience and Neuropsychology ***** Temporal clustering with spiking neurons and dynamic synapses: towards technological applications. J. Storck, F. Jakel, G. Deco ***** Mathematical and Computational Analysis ***** An evaluation of standard retrieval algorithms and a binary neural approach. V.J. Hodge, J. Austin Pseudo-outer product based fuzzy neural network fingerprint verification system. C. Quek, K. B. Tan, V. K. Sagar ***** Engineering and Design ***** A what-and-where fusion neural network for recognition and tracking of multiple radar emitters. E. Granger, M. A. Rubin, S. Grossberg, P. Lavoie ***** TECHNOLOGY & APPLICATIONS ***** Graph matching vs mutual information maximization for object detection. L.B. Shams, M. J. Brady, S. Schaal Naural networks for improved target differentiation and localization with sonar. B. Ayrulu, B. Barshan ------------------------------------------------------------------ Electronic access: www.elsevier.com/locate/neunet/. Individuals can look up instructions, aims & scope, see news, tables of contents, etc. Those who are at institutions which subscribe to Neural Networks get access to full article text as part of the institutional subscription. Sample copies can be requested for free and back issues can be ordered through the Elsevier customer support offices: nlinfo-f at elsevier.nl usinfo-f at elsevier.com or info at elsevier.co.jp ------------------------------ INNS/ENNS/JNNS Membership includes a subscription to Neural Networks: The International (INNS), European (ENNS), and Japanese (JNNS) Neural Network Societies are associations of scientists, engineers, students, and others seeking to learn about and advance the understanding of the modeling of behavioral and brain processes, and the application of neural modeling concepts to technological problems. Membership in any of the societies includes a subscription to Neural Networks, the official journal of the societies. Application forms should be sent to all the societies you want to apply to (for example, one as a member with subscription and the other one or two as a member without subscription). The JNNS does not accept credit cards or checks; to apply to the JNNS, send in the application form and wait for instructions about remitting payment. The ENNS accepts bank orders in Swedish Crowns (SEK) or credit cards. The INNS does not invoice for payment. ---------------------------------------------------------------------------- Membership Type INNS ENNS JNNS ---------------------------------------------------------------------------- membership with $80 or 660 SEK or Y 15,000 [including Neural Networks 2,000 entrance fee] or $55 (student) 460 SEK (student) Y 13,000 (student) [including 2,000 entrance fee] ----------------------------------------------------------------------------- membership without $30 200 SEK not available to Neural Networks non-students (subscribe through another society) Y 5,000 (student) [including 2,000 entrance fee] ----------------------------------------------------------------------------- Institutional rates $1132 2230 NLG Y 149,524 ----------------------------------------------------------------------------- Name: _____________________________________ Title: _____________________________________ Address: _____________________________________ _____________________________________ _____________________________________ Phone: _____________________________________ Fax: _____________________________________ Email: _____________________________________ Payment: [ ] Check or money order enclosed, payable to INNS or ENNS OR [ ] Charge my VISA or MasterCard card number ____________________________ expiration date ________________________ INNS Membership 19 Mantua Road Mount Royal NJ 08061 USA 856 423 0162 (phone) 856 423 3420 (fax) innshq at talley.com http://www.inns.org ENNS Membership University of Skovde P.O. Box 408 531 28 Skovde Sweden 46 500 44 83 37 (phone) 46 500 44 83 99 (fax) enns at ida.his.se http://www.his.se/ida/enns JNNS Membership c/o Professor Tsukada Faculty of Engineering Tamagawa University 6-1-1, Tamagawa Gakuen, Machida-city Tokyo 113-8656 Japan 81 42 739 8431 (phone) 81 42 739 8858 (fax) jnns at jnns.inf.eng.tamagawa.ac.jp http://jnns.inf.eng.tamagawa.ac.jp/home-j.html ----------------------------------------------------------------- From Dominique.Martinez at loria.fr Tue Mar 13 12:01:32 2001 From: Dominique.Martinez at loria.fr (Dominique Martinez) Date: Tue, 13 Mar 2001 18:01:32 +0100 (MET) Subject: Postdoctoral Fellowship in Neuromimetic Olfatory SEnsing Message-ID: Postdoctoral Fellowship in Neuromimetic Olfatory SEnsing (NOSE). A one year and a half postdoctoral fellowship is available immediately within the cooperative research project NOSE from INRIA. The first objective of this project will be to develop biologically inspired spiking neural network models of olfactory perception. The second objective will be to apply these models in an autonomous robot so as to mimic the animal behavior of tracking specific odours. The postdoctoral fellow will join the CORTEX group at LORIA-INRIA in Nancy, France. For further information see http://www.loria.fr/~rochel/nose or contact Dominique Martinez CORTEX group - LORIA Campus Scientifique - BP 239 54506 Vandoeuvre-Les-Nancy E-mail: dmartine at loria.fr Phone: (+33) 3-83-59-30-72 Fax: (+33) 3-83-41-30-79 From gomi at idea.brl.ntt.co.jp Tue Mar 13 20:36:54 2001 From: gomi at idea.brl.ntt.co.jp (Hiroaki GOMI) Date: Wed, 14 Mar 2001 10:36:54 +0900 Subject: RESEARCH POSITIONS AVAILABLE at NTT Communication Sci. Labs. Message-ID: <20010314103654Z.gomi@idea.brl.ntt.co.jp> RESEARCH POSITIONS AVAILABLE NTT (Nippon Telegraph and Telephone Corporation) Communication Science Labs. Research associate (post doctoral fellow) and research specialist positions are currently available in the human speech & motor control group at NTT Communication Science Laboratories. One research topic focuses on the human motor control mechanisms in the mechanical interaction with environments, and its brain information processing. The candidate should have a strong background in human motor control studies. Some skills of C programming for controlling experimental system and Matlab for data analyses are desirable but not required. Candidates with training in psychological approaches to motor control are particularly encouraged to apply. Another topic focuses on information processing for speech and language. The candidate should have a strong background in the fields of speech and language. The position requires strong programming skills in C and/or C++ with experience in programming mathematical algorithms, and knowledge of signal processing and machine learning. The salary will be decided according to company regulation. The research center for these topics is located in Atsugi near Tokyo, Japan (see our web page). The positions are tenable for one year and are renewable every year until the limit of two years. Please send curriculum vitae including research and programming experiences, names and contact details of referees, and representative publications to the address listed below. Curriculum vitae should be sent until March 23, 2001. Informal inquiries should be set to the e-mail address. E-mail inquiries and applications are encouraged. We prefer a starting date of April 1, 2001 but this is open to negotiation. Application should be sent to: Dr. Eisaku Maeda soukatsu at cslab.kecl.ntt.co.jp Senior Manager of NTT Communication Science Laboratories Tel: +81-774-93-5040 Fax: +81-774-93-5045 address: NTT Communication Science Labs. Hikaridai 2-4, Seika-cho, Soraku-gun, Kyoto-pref. 619-0237, Japan web page: http://www.kecl.ntt.co.jp From swatanab at pi.titech.ac.jp Wed Mar 14 20:44:29 2001 From: swatanab at pi.titech.ac.jp (Sumio Watanabe) Date: Thu, 15 Mar 2001 10:44:29 +0900 Subject: CFP: Special Session in KES'2001: Geometry and Statistics in NN Message-ID: <004f01c0acf1$79fd19a0$918a7083@deskpowerts> Dear Connectionists, We have a special session on Geometry and Statistics in NN in the international conference KES'2001, which will be held in Oska and Nara in Japan, 6,7,8, September, 2001. http://www.bton.ac.uk/kes/kes2001/ In the session "Geometry and Statistics in Neural Network Learning Theory", we study the statistical problem caused by non-identifiability of layered learning machines. (See the followings). We would like to invite some researchers who will take part in this session and present a paper. http://www.bton.ac.uk/kes/kes2001/sessions.html. The authors who will be invited: Shun-Ichi Amari (RIKEN Brain Science Institute,Japan) Kenji Fukumizu (Insitute of Statistical Mathematics,Japan) Katsuyuki Hagiwara (Mie University,Japan) Sumio Watanabe(Tokyo Institute of Technology,Japan) If you are interested in this special session, please contact the following e-mail address by March 31th, 2001. swatanab at pi.titech.ac.jp Thank you very much. Sincerely, Dr. Sumio Watanabe, Associate Professor Advanced Information Processing Division Precision and Intelligence Laboratory Tokyo Institute of Technology (Fax) +81-45-924-5018 E-mail: swatanab at pi.titech.ac.jp http://watanabe-www.pi.titech.ac.jp/~swatanab/index.html ***************Special Session *************** Geometry and Statistics in Neural Network Learning Theory Non-identifiability: A learning machine p(x|w) with a parameter w is called identifiable if the mapping from w to p(x|w) is one-to-one. It should be emphasized that almost all learning machines with hidden parts are not identifiable, resulting that the manifolds of parameters have singular Fisher metrics. Problem: If a non-identifiable learning machine can almost approximate the true distribution, then, because of the finiteness of training samples, it is often in an almost redundant state. In such a case, neither the distribution of the MLE nor the Bayes a posteriori distrbution is subject to the asymptotically normal distribution. The conventional statistical asymptotic theory can not be applied to analyze the learning curves. Purpose: We would like to clarify the relation between the learning curve and geometry of neuro-manifolds when the set of true parameters is an analytic set (the set defined as zeros of an analytic function). Methods: The following topics will be discussed. (1) Information Geometry of Singular Neuro-Manifold. (2) Theory of Order Statistic. (3) Weak Convergence and Empirical Prcess. (4) Conic Singularities. (5) Algebraic Geometry and Resolution of Singuralities. (6) Zeta function of Kullback Information and Prior. ******************************************* From Pete.Roper at nih.gov Wed Mar 14 11:24:48 2001 From: Pete.Roper at nih.gov (Pete Roper) Date: Wed, 14 Mar 2001 11:24:48 -0500 Subject: Postdoctoral Fellow Position Available Message-ID: <3AAF9B4E.438C@nih.gov> Postdoctoral Fellow Vacancy in the Laboratory of Cellular and Synaptic Neurophysiology; National Institute of Child Health and Human Development A position is available for a postdoctoral fellow within the Laboratory of Dr Chris J. McBain, Laboratory of Cellular and Synaptic Neurophysiology at the NICHD. Candidates should have experience in computational neurosciences with particular emphasis on the modeling of cortical neuronal networks. The candidate will work with an outstanding team of electrophysiologists to model excitatory and inhibitory synaptic circuitry and voltage gated potassium channels expressed on local circuit inhibitory neurons and principal neurons within the CA3 subfield of the hippocampal formation (see Atzori and McBain, Nature Neurosci. 3:791, 2000; Toth et al JNeurosci 15: 8279; Toth and McBain Nature Neurosci 1; 572, 1998). Candidates should have less than five years postdoctoral experience. Interested applicants should email Dr Chris McBain at chrismcb at codon.nih.gov -- ------------------------------ Dr P Roper Mailing Address: Postdoctoral Fellow BSA Building, Suite 350 Mathematical Research Branch, 9190 Rockville Pike National Institute of Diabetes and Bethesda, MD 20892-2690 Digestive and Kidney Diseases, USA National Institutes of Health Bethesda, MD 20892. E?mail: pete.roper at nih.gov Phone: (301) 496-9644 Fax: (301) 402-0535 From mike at stats.gla.ac.uk Wed Mar 14 09:38:09 2001 From: mike at stats.gla.ac.uk (Mike Titterington) Date: Wed, 14 Mar 2001 14:38:09 +0000 (GMT) Subject: JOB : Postdoctoral research assistant at Glasgow Message-ID: A POSTDOCTORAL RESEARCH OPPORTUNITY AT THE UNIVERSITY OF GLASGOW (Ref 129/01) APPROXIMATE APPROACHES TO LIKELIHOOD AND BAYESIAN STATISTICAL INFERENCE IN INCOMPLETE-DATA PROBLEMS Applications are sought for a Post-Doctoral Research Assistantship (IA) at the University of Glasgow. The appointee will be based in the Department of Statistics, under the direction of Professor D. M. Titterington. The appointment will be made on the standard IA scale of 16775 - 25213 and will be for up to 36 months, starting on October 1, 2001, or as soon as possible thereafter. The post is funded by the UK Engineering and Physical Sciences Research Council (EPSRC). The research concerns an area in which algorithms such as the EM algorithm and Markov chain Monte Carlo procedures can suffer from high computational complexity and the appointee will investigate and extend variational and other approximations that are currently available in the statistical and neural-computing literatures. Applications, supported by full curriculum vitae and the names of three referees, should be sent to Professor D. M. Titterington, Department of Statistics, University of Glasgow, Glasgow G12 8QQ, Scotland, to arrive not later than April 13, 2001. Informal enquiries and requests for further particulars can be made to mike at stats.gla.ac.uk. ================================= D.M. Titterington, Department of Statistics, University of Glasgow, Glasgow G12 8QQ, Scotland, UK. mike at stats.gla.ac.uk Tel (44)-141-330-5022 Fax (44)-141-330-4814 http://www.stats.gla.ac.uk/~mike From inmanh at cogs.susx.ac.uk Thu Mar 15 11:17:53 2001 From: inmanh at cogs.susx.ac.uk (Inman Harvey) Date: Thu, 15 Mar 2001 16:17:53 +0000 Subject: MSc in Evolutionary and Adaptive Systems Message-ID: <3AB0EB31.9653C7EA@cogs.susx.ac.uk> The Evolutionary and Adaptive Systems (EASy) group at the University of Sussex is probably the largest such multidisciplinary research group in the world, working on a wide range of topics where Computer Science and Complex Systems and AI and Artificial Life swap ideas with Biology. We have over 30 active researchers at doctoral and postdoctoral level, plus a similar number of Masters students. The EASy MSc is a one year course with 2 terms of coursework followed by a major supervised project in a relevant area. If you are interested you can find out about us on http://www.cogs.susx.ac.uk/lab/adapt/index.html **Expanded funding** Due to its success over the last 4 years the UK EPSRC has now dramatically increased the financial support for this course. The number of studentships available has increased very significantly, awarded competitively: UK students tuition + living expenses, other EU students tuition fees only. Other international and local students not awarded studentships welcome on self-funding basis, part-time option over 2 years also available. The expanded funding also includes new facilities and new courses to be added. Faculty directly involved in the course include Dr Inman Harvey - artificial evolution, evolutionary robotics, artificial life Dr Phil Husbands - evolutionary computation, GasNets for robotics Dr Ezequiel Di Paolo - evolving collective behaviour, homeostasis, autopoiesis Dr Adrian Thompson - evolvable hardware, evolutionary electronics Other faculty here at Sussex in associated areas include Prof John Maynard Smith (Evolutionary theory) Prof Maggie Boden (Philosophy, Creativity, Artificial Life) Prof Andy Clark (Philosophy) Prof Tom Collett (Ant and bee navigational behaviour) Prof Mick O'Shea (Neuroscience) For further information and applications contact Linda Thompson COGS University of Sussex Brighton BN1 9QH pgadmissions at cogs.susx.ac.uk http://www.cogs.susx.ac.uk/lab/adapt/index.html =============================== -- Inman Harvey >> Evolutionary and Adaptive Systems Group << >> COGS, Univ. of Sussex, Brighton BN1 9QH, UK << inmanh at cogs.susx.ac.uk >> http://www.cogs.susx.ac.uk/users/inmanh/ << From apbraga at cpdee.ufmg.br Fri Mar 16 12:52:08 2001 From: apbraga at cpdee.ufmg.br (Antonio de Padua Braga) Date: Fri, 16 Mar 2001 14:52:08 -0300 Subject: Multi-objective optimization/bias-variance Message-ID: <3AB252C8.AFC7FDC5@cpdee.ufmg.br> Dear Connectionists, The following paper has just been published by Neurocomputing. The idea of the paper is to balance the error of the training set and the norm of the weight vectors with a multi-objective optimization approach to avoid over-fitting. Copies are available on request. We apologize in advance for any multiple postings that may be received. *********************************************************************** Improving generalization of MLPs with multi-objective optimization Teixeira, R.A., Braga, A.P., Takahashi, R.H.C. And Rezende, R. Neurocomputing. Volume 35, pages 189-194. ABSTRACT This paper presents a new learning scheme for improving generalization of Multilayer Perceptrons (MLPs). The algorithm uses a multi-objective optimization approach to balance between the error of the training data and the norm of network weight vectors to avoid over-fitting. The results are compared with Support Vector Machines (SVMs) and standard backpropagation. *********************************************************************** -- Prof. Antonio de Padua Braga, Depto. Eng. Eletronica, Campus da UFMG (Pampulha), C. P. 209, 30.161-970, Belo Horizonte, MG, Brazil Tel:+55 31 4994869, Fax:+55 31 4994850, Email:apbraga at cpdee.ufmg.br, http://www.cpdee.ufmg.br/~apbraga From lunga at ifi.unizh.ch Sun Mar 18 12:34:22 2001 From: lunga at ifi.unizh.ch (Max Lungarella) Date: 18 Mar 2001 18:34:22 +0100 Subject: EDEC2001 (2nd call) Message-ID: <3AB4F19E.C10C2391@ifi.unizh.ch> Note: Our apologies if you receive this posting more than once. CALL FOR CONTRIBUTIONS EDEC2001 - EMERGENCE AND DEVELOPMENT OF EMBODIED COGNITION (2nd call) Symposium at the 3rd International Conference on Cognitive Science August 27-31, 2001, Beijing, China SCOPE The objective of the symposium is to bring together researchers from cognitive science, psychology, engineering, robotics, artificial intelligence, philosophy, and related fields so as to further our understanding of embodiment and development, in particular their mutual relationship. Ultimately, the goal is to understand the emergence of high-level cognition of an organism interacting with its physical and social environment over extended periods of time. FOCUS The symposium will focus on research that explicitly takes embodiment into account, either at the level of computational models, or real-world devices, and on empirical work that explicitly attempts to explain the relation of developmental processes to embodiment. Finally, contributions giving a broad and novel philosophical or methodological view on embodied cognition are welcome. CONTRIBUTIONS Contributions are solicited from the following areas (but not restricted to this list): - Cognitive developmental robotics - Neural mechanisms of learning and development (e.g. neural networks, statistical, information theoretic) - Development of sensory and motor systems - Perception-action coupling, sensory-motor coordination - Categorization, object exploration - Communication and Social interaction - Methodologies - Debates and philosophical issues (e.g. constructivism vs. selectionism, nature nurture, scalability, symbol grounding) ORGANIZATION This will be a one-day symposium with a number of talks, with a lot of room for discussion, and a poster session. The poster session will be over a cocktail to ensure relaxed atmosphere. FORM OF CONTRIBUTION Contributions can be in the form of full papers, or abstracts for posters. Full papers must be 5 pages maximum, with fonts at least 12 pt. For posters and work in progress, please submit a one-page abstract. Accepted contributions will be published in the proceedings of the ICCS 2001. GUIDELINES FOR ABSTRACT/PAPER SUBMISSION The contribution should be submitted electronically to lunga at ifi.unizh.ch (Max Lungarella) in pdf or MS Word files. IMPORTANT DATES Submission deadline (full-length paper and abstract): April 30, 2001 Notice of acceptance: May 31, 2001 PROGRAM COMMITTEE Rolf Pfeifer (AI Lab, University of Zurich, chair) Max Lungarella (AI Lab, University of Zurich, Switzerland) Yasuo Kuniyoshi (Electrotechnical Laboratory, Tsukuba, Japan) Olaf Sporns (Department of Psychology, Indiana University, Bloomington, IN, USA) Giorgio Metta (LIRA-Lab, University of Genova, Italy) Giulio Sandini (LIRA-Lab, University of Genova, Italy) Rafael Nunez (University of Fribourgh, Switzerland, and University of California, Berkeley) From geoff at giccs.georgetown.edu Sun Mar 18 22:09:01 2001 From: geoff at giccs.georgetown.edu (Geoff Goodhill) Date: Sun, 18 Mar 2001 22:09:01 -0500 Subject: Papers available: topographic mappings Message-ID: <200103190309.WAA30535@brecker.giccs.georgetown.edu> The following 3 recent papers about topographic mappings (both natural and artificial) can now be downloaded from www.giccs.georgetown.edu/~geoff/pubs.html Haese, K. & Goodhill, G.J. (2001). Auto-SOM: Recursive Parameter Estimation for Guidance of Self-Organizing Feature Maps. Neural Computation , 13, 595-619. Goodhill, G.J. & Cimponeriu, A. (2000). Analysis of the elastic net model applied to the formation of ocular dominance and orientation columns. Network, 11, 153-168. Goodhill, G.J. (2000). Dating behavior of the retinal ganglion cell. Neuron, 25, 501-503. Regards, Geoff Geoffrey J Goodhill, PhD Assistant Professor, Department of Neuroscience & Georgetown Institute for Cognitive and Computational Sciences Georgetown University Medical Center 3900 Reservoir Road NW, Washington DC 20007 Tel: (202) 687 6889, Fax: (202) 687 0617 Email: geoff at giccs.georgetown.edu Homepage: www.giccs.georgetown.edu/labs/cns From ahirose at info-dev.rcast.u-tokyo.ac.jp Mon Mar 19 02:13:32 2001 From: ahirose at info-dev.rcast.u-tokyo.ac.jp (ahirose@info-dev.rcast.u-tokyo.ac.jp) Date: Mon, 19 Mar 2001 16:13:32 +0900 (JST) Subject: CFP KES'2001 Special Session: Complex-valued neural networks Message-ID: <200103190713.QAA08147@info-dev.rcast.u-tokyo.ac.jp> Dear Connectionists, The Special Session "Complex-valued neural networks and the applications" is now being organized for the International Conference KES'2001 Osaka, September 6-8, 2001 (http://www.bton.ac.uk/kes/kes2001/) In these years the complex-valued networks expand the application fields in optoelectronic imaging, remote sensing, quantum neural devices and systems, spatiotemporal analysis of physiological neural systems as well as artificial neural processing. The potentially wide applicability yields new aspects of theories required for novel or more effective functions and mechanisms. The Special Session aims at an inspiring discussion on the recent progress and the future development. A few more speakers will be invited in addition to: Hiroyuki Aoki (Tokyo National College of Technology) George Georgiou (California State University) Akira Hirose (Tokyo University) Iku Nemoto (Tokyo Denki University) Tohru Nitta (Electrotechnical Laboratory) If you are interested, please make contact by March 31 with the Session Chair: Akira Hirose, University of Tokyo ahirose at ee.t.u-tokyo.ac.jp -- From tomsimpson at hotmail.com Mon Mar 19 13:31:08 2001 From: tomsimpson at hotmail.com (Tom Simpson) Date: Mon, 19 Mar 2001 18:31:08 -0000 Subject: New Issue of Connexions Message-ID: APOLOGIES FOR CROSS-POSTING This is to announce the release of a new issue of Connexions - an online journal of issues in Philosophy and Cognitive Science. In this issue we have two papers: 1. 'Connectionism is Nothing but Control Theory' by Asim Roy Abstract: The paper shows that connectionist systems are based on standard control theory notions. It does so by first examining the notion of a controller in any system and then establishing that many of the simpler connectionist learning methods - like back-propagation, adaptive resonance theory (ART), reduced coulomb energy (RCE), and radial basis function (RBF) - use controllers in them. The paper shows the existence of controllers in these methods and shows that these controllers may be (1) within the learning system itself, (2) outside of the learning system or (3) a combination of the two. By logical extension, more complex connectionist systems, ones that use these simpler learning mechanisms within them, are also in turn using controllers and are therefore based on control theoretic concepts. The analysis of these connectionist systems is performed purely on a logical basis, by a logical analysis of their conceptual structure, and has nothing to do with their implementation, whether by use of neurocomputers or other kinds of computers. In general, this analysis implies that control theoretic notions are applicable to developing systems similar to the brain and refutes the claim in connectionism that their methods do not embody standard control theory concepts, that they have introduced a qualitatively new set of concepts and mechanisms and 2. 'On the Origin of Symbols' by Thomas E. Dickins. Abstract: Through a synthesis of Dunbars (1993, 1996) and Gomezs (1998a, b) hypotheses on the evolution of language a further hypothesis about the origins of symbolic communication is made that relies upon simple learning. The aim of this speculation is to propose a specific origin story for symbolic communication and to marry behaviourist and cognitivist concerns about language. It is argued that this order of approach will enforce a more realistic parsimony on future models of language evolution. We also have a review, by Tom Stafford, of 'Darwin Among the Machines' by George Dyson. Connexions is run by the Department of Philosophy at the University of Sheffield, England, and can be found at http://www.shef.ac.uk/uni/academic/N-Q/phil/connex/ Connexions is an on-line journal for all issues in the Philosophy of Cognitive Science. We primarily publish work-in-progress with the intention of exposing this work to friendly but rigorous criticism, and, more broadly, of generating debate on issues within cognitive science. Comments and replies in response to work published in the journal are encouraged, and can be mailed to the editor or posted on our discussion list Connex-L. Contributions for subsequent publications are also welcome. Details on how to join the list, post messages and submit articles and reviews can be found on the site, together with previous issues and further information about Connexions. Thank you for your time. Tom Simpson Editor tomsimpson at hotmail.com ----------------------------------------------------------- Tom Simpson Department of Philosophy University of Sheffield Sheffield S10 2TN http://www.shef.ac.uk/misc/personal/pip98ts _________________________________________________________________________ Get Your Private, Free E-mail from MSN Hotmail at http://www.hotmail.com. From a.burkitt at medoto.unimelb.edu.au Tue Mar 20 21:59:08 2001 From: a.burkitt at medoto.unimelb.edu.au (Anthony BURKITT) Date: Wed, 21 Mar 2001 13:59:08 +1100 Subject: Three preprints on integrate-and-fire neuron models Message-ID: <2DE39EB1B64BD31184C100C00D006EF9A16A38@mail.medoto.unimelb.edu.au> Dear Connectionists, I would like to announce the availability of three papers that have been accepted for publication and are of potential interest to people working on integrate-and-fire neurons. Electronic copies are available at the site: http://www.medoto.unimelb.edu.au/people/burkitta/Pubs.html or contact me at a.burkitt at medoto.unimelb.edu.au Feedback and comments are naturally very welcome. Cheers, Tony Burkitt =============================================== Synchronization of the neural response to noisy periodic synaptic input A. N. Burkitt and G. M. Clark The timing information contained in the response of a neuron to noisy periodic synaptic input is analyzed for the leaky integrate-and-fire neural model. We address the question of the relationship between the timing of the synaptic inputs and the output spikes. This requires an analysis of the interspike interval distribution of the output spikes, which is obtained in the Gaussian approximation. The conditional output spike density in response to noisy periodic input is evaluated as a function of the initial phase of the inputs. This enables the phase transition matrix to be calculated, which relates the phase at which the output spike is generated to the initial phase of the inputs. The interspike interval histogram and the period histogram for the neural response to ongoing periodic input are then evaluated by using the leading eigenvector of this phase transition matrix. The synchronization index of the output spikes is found to increase sharply as the inputs become synchronized. This enhancement of synchronization is most pronounced for large numbers of inputs and lower frequencies of modulation, and also for rates of input near the critical input rate. However, the mutual information between the input phase of the stimulus and the timing of output spikes is found to decrease at low input rates as the number of inputs increases. The results show close agreement with those obtained from numerical simulations for large numbers of inputs. http://www.medoto.unimelb.edu.au/people/burkitta/periodic.ps.zip Accepted for publication in Neural Computation (to appear). =============================================== Shot noise in the leaky integrate-and-fire neuron N. Hohn and A. N. Burkitt We study the influence of noise on the transmission of temporal information by a leaky integrate-and-fire neuron using the theory of shot noise. The model includes a finite number of synapses and has a membrane potential variance de facto modulated by the input signal. The phenomenon of stochastic resonance in spiking neurons is analytically exhibited using an inhomogeneous Poisson process model of the spike trains, and links with the traditional Ornstein-Uhlenbeck process obtained by a diffusion approximation are given. It is shown that the modulated membrane potential variance inherent to the model gives better signal processing capabilities than the diffusion approximation. http://www.medoto.unimelb.edu.au/people/burkitta/article_PRE_ES7178.ps.zip Published in Phys. Rev. E 63, 031902. =============================================== Balanced neurons: Analysis of leaky integrate-and-fire neurons with reversal potentials A. N. Burkitt A new technique is presented for analyzing leaky integrate-and-fire neurons that incorporates reversal potentials, which impose a biologically realistic lower bound to the membrane potential. The time distribution of the synaptic inputs is modeled as a Poisson process. The analysis is carried out in the Gaussian approximation, which comparison with numerical simulations confirms is most accurate in the limit of a large number of inputs. The hypothesis that the observed variability in the spike times of cortical neurons is caused by a balance of excitatory and inhibitory synaptic inputs is supported by the results for the coefficient of variation of the interspike intervals. It's value decreases with both increasing numbers and amplitude of inputs, and is consistently lower than 1.0 over a wide range of realistic parameter values. The dependence of the output spike rate upon the rate, number and amplitude of the synaptic inputs, as well as upon the value of the inhibitory reversal potential, are given. http://www.medoto.unimelb.edu.au/people/burkitta/balanced.ps.zip Accepted for publication in Biological Cybernetics (to appear). ====================ooOOOoo==================== Anthony N. Burkitt The Bionic Ear Institute 384-388 Albert Street East Melbourne, VIC 3002 Australia Email: a.burkitt at medoto.unimelb.edu.au http://www.medoto.unimelb.edu.au/people/burkitta Phone: +61 - 3 - 9283 7510 Fax: +61 - 3 - 9283 7518 =====================ooOOOoo=================== From bengio at idiap.ch Wed Mar 21 05:26:42 2001 From: bengio at idiap.ch (Samy Bengio) Date: Wed, 21 Mar 2001 11:26:42 +0100 (MET) Subject: Several open positions in speech, vision, machine learning at IDIAP Message-ID: SEVERAL OPEN POSITIONS IN SPEECH, COMPUTER VISION, MACHINE LEARNING, AND MULTIMODAL INTERFACES The Dalle Molle Institute for Perceptual Artificial Intelligence (IDIAP, http://www.idiap.ch) is a semi-private research institute affiliated with the Swiss Federal Institute of Technology at Lausanne (EPFL) and the University of Geneva. Located in Martigny (Valais, CH), IDIAP is partly funded by the Swiss Federal Government, the State of Valais, and the City of Martigny, and is involved in numerous national and international (European) projects. IDIAP is mainly carrying research and development in the fields of speech and speaker recognition, computer vision, and machine learning, and is recognized as a university level research laboratory, involving permanent senior scientists, postdocs, and PhD students (usually awarded an EPFL degree). IDIAP currently numbers around 35-40 scientists. IDIAP, in close collaboration with EPFL (mainly with the Signal Processing Laboratory of Prof. Murat Kunt, http://ltswww.epfl.ch) will soon be the "Leading House" of a large Research Network (National Center of Competence in Research) on "Interactive Multimodal Information Management", which also opens up several new opportunities for long term positions at different levels. In view of the resulting (present and future) growth of the Institute, IDIAP currently welcomes applications of talented candidates at all levels with expertise or strong interest in the fields of speech processing, computer vision, machine learning, and multimodal interaction. The open positions include: management and senior positions (including one scientific deputy director and one speech processing group leader), project leaders, postdocs, and PhD students. Two EPFL tenure track positions at the Assistant Professor (with most of the research responsibilities located at IDIAP) are also available. Preference will be given to candidates with experience in one or several of the following areas: signal processing, statistical pattern recognition (typically applied to speech and scene analysis), neural networks, hidden Markov models, speech and speaker recognition, computer vision, human/computer interaction (dialog). Senior and postdoc candidates should also have a proven record of high quality research and publications. All applicants should be experienced in C/C++ programming and familiar with the Unix environment; they should also be able to speak and write in English (and be willing to learn French). LOCATION: IDIAP is located in the town of Martigny (http://www.martigny.ch) in Valais, a scenic region in the South of Switzerland, surrounded by the highest mountains of Europe, and offering exciting recreational activities (including hiking, climbing and skiing), as well as varied cultural activities. It is also within close proximity to Montreux, Lausanne (EPFL) and Lake Geneva, and centrally located for travel to other parts of Europe. PROSPECTIVE CANDIDATES should send their detailed CV, together with a motivation letter and 3 reference letters, to: Prof. Herv? Bourlard, Director of IDIAP, P.O. Box 592, Simplon, 4 CH-1920 Martigny Switzerland Email: bourlard at idiap.ch Phone: +41-27-721.77.20; Fax: +41-27-721.77.12 ----- Samy Bengio Research Director. Machine Learning Group Leader. IDIAP, CP 592, rue du Simplon 4, 1920 Martigny, Switzerland. tel: +41 27 721 77 39, fax: +41 27 721 77 12. mailto:bengio at idiap.ch, http://www.idiap.ch/~bengio From J.Hancock at cs.rhul.ac.uk Wed Mar 21 07:58:13 2001 From: J.Hancock at cs.rhul.ac.uk (John Hancock) Date: Wed, 21 Mar 2001 12:58:13 +0000 Subject: Bioinformatics Postdoc, Univerity of London UK Message-ID: Postdoctoral Research Assistant Bioinformatics Research Group, Department of Computer Science, Royal Holloway University of London We are looking for a highly motivated individual to join the bioinformatics research group in the Department. The aim of the project, funded by the BBSRC's bioinformatics initiative, is to apply machine learning techniques (such as the Support Vector Machine) to the identification of coding regions, gene promoters and other sequence features in genomic sequences, with special emphasis on plant genomes. This is a three year appointment commencing as soon as possible. The Department of Computer Science at Royal Holloway has a leading position in the study of theory and practice of machine learning and in particular the development of the Support Vector learning technique and other kernel-based techniques. The Department is located on Royal Holloway's pleasant campus (see http://www.rhul.ac.uk/) in Englefield Green, Surrey, close to London. The successful candidate will work with John Hancock, Alex Gammerman and Victor Solovyev in collaboration with Peter Bramley in the School of Biological Sciences. We are looking for someone with a strong background in computational biology, particularly as applied to genomic sequence analysis, and preferably also some programming skills and knowledge of machine learning. The post would also be ideal for someone with a computer science or maths background who is looking to move into bioinformatics. Salary is in the range 18,909 to 22,599 UK pounds per annum inclusive of London Allowance. Information on the Department may be found at www.cs.rhul.ac.uk/. Informal enquiries may be directed to John Hancock (J.Hancock at cs.rhul.ac.uk). See http://www.cs.rhul.ac.uk/home/jhancock/ for more information on the project and John Hancock's research. Further details and an application form can be obtained from The Personnel Office, Royal Holloway, University of London, Egham, Surrey TW20 0EX; fax: +44 (0)1784 473527; tel: +44 (0)1784 414241; email Sue.Clarke at rhul.ac.uk Please quote reference KB/1717. The closing date for applications is 23rd April 2001. We positively welcome applications from all sections of the community. *************************************************** Dr John M. Hancock Reader in Computational Biology, Director of Graduate Studies, Department of Computer Science, Royal Holloway University of London, Egham, Surrey TW20 0EX, U.K. *************************************************** Phone: +44 (01784) 414 256 Fax: +44 (01784) 439 786 Mobile: +44 (077) 9046 6302 E-mail: J.Hancock at cs.rhul.ac.uk WWW: http://www.cs.rhul.ac.uk/home/jhancock/ *************************************************** From ishii at is.aist-nara.ac.jp Thu Mar 22 01:22:02 2001 From: ishii at is.aist-nara.ac.jp (Shin Ishii) Date: Thu, 22 Mar 2001 15:22:02 +0900 Subject: Preprint available Message-ID: <200103220622.PAA16405@dec127.aist-nara.ac.jp> Dear connectionists, We are pleased to inform you that the following preprint is now available on our Web page: http://www.aist-nara.ac.jp/~ishii/Papers/nc_psn.ps.gz Comments and suggestion would be greatly appreciated. Authors: Ken-ichi Amemori and Shin Ishii Title: Gaussian process approach to spiking neurons for inhomogeneous Poisson inputs (Neural Computation, to appear) Keywords: spiking neuron, stochastic process, Markov-Gaussian process, inhomogeneous Poisson input, first passage time density Abstract: This article presents a new theoretical framework to consider the dynamics of a stochastic spiking neuron model with general membrane response to input spike. We assume that the input spikes obey an inhomogeneous Poisson process. The stochastic process of the membrane potential then becomes a Gaussian process. When a general type of the membrane response is assumed, the stochastic process becomes a Markov-Gaussian process. We present a calculation method for the membrane potential density and the firing probability density. Our new formulation is the extension of the existing formulation based on diffusion approximation. Although the single Markov assumption of the diffusion approximation simplifies the stochastic process analysis, the calculation is inaccurate when the stochastic process involves a multiple Markov property. We find that the variation of the shape of the membrane response, which has often been ignored in existing stochastic process studies, significantly affects the firing probability. Our approach can consider the reset effect, which has been difficult to be dealt with by analysis based on the first passage time density. --- Shin Ishii Nara Institute of Science and Technology http://www.aist-nara.ac.jp/~ishii/ From erik at bbf.uia.ac.be Thu Mar 22 10:09:44 2001 From: erik at bbf.uia.ac.be (Erik De Schutter) Date: Thu, 22 Mar 2001 16:09:44 +0100 Subject: European short-term fellowships in neuroinformatics and computational neuroscience Message-ID: The EU Thematic Network Computational Neuroscience and Neuroinformatics offers short-term fellowships to nationals of the EU and associated countries for visits to laboratories in the EU or associated countries. Fellowships cover travel and accommodation costs for a visit of 2 to 12 weeks. Applications are evaluated once every month on a competitive basis. Fellowships can be awarded within three months of the application. Award per fellowship varies between 600 to 2000 =A4. More information and electronic application is available at http:www.neuroinf.org Prof. E. De Schutter University of Antwerp, Belgium coordinator of the Thematic Network Computational Neuroscience and Neuroinformatics http://www.bbf.uia.ac.be From john at cs.rhul.ac.uk Fri Mar 23 12:29:22 2001 From: john at cs.rhul.ac.uk (John Shawe-Taylor) Date: Fri, 23 Mar 2001 17:29:22 +0000 (GMT) Subject: Applications of learning to text and images Message-ID: Expressions of interest are invited for the following workshop: NeuroCOLT/KerMIT workshop on Applications of Learning to Text and Images Windsor, 30 April - 2 May 2001 Cumberland Lodge, Windsor, UK http://www.neurocolt.org/cumberland2001.html REGISTRATION DEADLINE: APRIL 3rd The continuous growth of information repositories in digital format (eg the World Wide Web) makes it necessary to develop entirely new - and automatic - ways of accessing it. Learning methods are one of the most promising research directions. This workshop will bring together experts from information retrieval, text categorization, machine vision, and learning theory. The aim is to provide an understanding of the state of the art, and of the open problems to which learning technology could be applied. The list of invited speakers is on the web site - they will be giving two hour tutorial style talks. For registration and constantly updated information, pls click here: http://www.neurocolt.org/cumberland2001.html The cost will be approximately 100 UKP per day inclusive of bed and full board. ************************************************************** John Shawe-Taylor J.Shawe-Taylor at cs.rhul.ac.uk Dept of Computer Science, Royal Holloway, University of London Phone: +44 1784 443430 Fax: +44 1784 439786 ************************************************************** From shultz at psych.mcgill.ca Fri Mar 23 10:55:53 2001 From: shultz at psych.mcgill.ca (Thomas R. Shultz) Date: Fri, 23 Mar 2001 10:55:53 -0500 Subject: recent papers on cognitive and language development Message-ID: <4.3.1.0.20010323104943.00a7cc30@127.0.0.1> Recent papers on cognitive and language development that may be of interest to readers of this list Shultz, T. R., & Bale, A. C. (2001, in press). Neural network simulation of infant familiarization to artificial sentences: Rule-like behavior without explicit rules and variables. Infancy. A fundamental issue in cognitive science is whether human cognitive processing is better explained by symbolic rules or by sub-symbolic neural networks. A recent study of infant familiarization to sentences in an artificial language claims to have produced data that can only be explained by symbolic rule learning and not by unstructured neural networks. Here we present successful unstructured neural network simulations of the infant data, showing that these data do not uniquely support a rule-based account. In contrast to other simulations of these data, the present simulations cover more aspects of the data with fewer assumptions about prior knowledge and training, using a more realistic coding scheme based on sonority of phonemes. The networks show exponential decreases in attention to a repeated sentence pattern, more recovery to novel sentences inconsistent with the familiar pattern than to novel sentences consistent with the familiar pattern, occasional familiarity preferences, more recovery to consistent novel sentences than to familiarized sentences, and extrapolative generalization outside the range of the training patterns. A variety of predictions suggest the utility of the model in guiding future psychological work. The evidence, from these and other simulations, supports the view that unstructured neural networks can account for the existing infant data. ================ Sirois, S., Buckingham, D., & Shultz, T. R. (2000). Artificial grammar learning by infants: An auto-associator perspective. Developmental Science, 4, 442-456. This paper reviews a recent article suggesting infants use a system of algebraic rules to learn an artificial grammar (Marcus, Vijayan, Bandi Rao, & Vishton, 1999). In three reported experiments, infants exhibited increased responding to auditory strings that violated the pattern of elements they were habituated to. We argue that a perceptual interpretation is more parsimonious, as well as more consistent with a broad array of habituation data. We report successful neural network simulations that implement a lower-level interpretation and capture the empirical regularities reported by Marcus and colleagues (1999). The discussion puts the simulation results in the context of the broader debate about interpreting infant habituation. Other neural network models of habituation in general, and of the Marcus et al. (1999) task specifically, are discussed. ================ Buckingham, D., & Shultz, T. R. (2000). The developmental course of distance, time, and velocity concepts: A generative connectionist model. Journal of Cognition and Development, 1, 305-345. Connectionist simulations of children's acquisition of distance (d), time (t), and velocity (v) concepts using a generative algorithm, cascade-correlation, are reported. Rules that correlated most highly with network responses during training were consistent with the developmental course of children's concepts. Networks integrated the defining dimensions of the concepts first by identity rules (e.g., v = d), then additive rules (e.g., v = d t), and finally multiplicative rules (e.g., v = d / t). The results are discussed in terms of similarity to children's development, the contribution of connectionism to the study of cognitive development, contrasts with alternate models, and directions for future research. ================ Oshima-Takane, Y., Takane, Y., & Shultz, T. R. (1999). The learning of first and second pronouns in English: Network models and analysis. Journal of Child Language, 26, 545-575. Although most English-speaking children master the correct use of first and second person pronouns by three years, some children show persistent reversal errors in which they refer to themselves as you and to others as me. Recently, such differences have been attributed to the relative availability of overheard speech during the learning process. The present study tested this proposal with feed-forward neural networks learning these pronouns. Network learning speed and analysis of their knowledge representations confirmed the importance of exposure to shifting reference provided by overheard speech. Errorless pronoun learning was linked to the amount of overheard speech, interactions with a greater number of speakers, and prior knowledge of the basic-level kind PERSON. Preprints and reprints can be found at http://www.psych.mcgill.ca/perpg/fac/shultz/default.htm Cheers, Tom -------------------------------------------------------- Thomas R. Shultz, Professor, Department of Psychology, McGill University, 1205 Penfield Ave., Montreal, Quebec, Canada H3A 1B1. E-mail: shultz at psych.mcgill.ca Updated 23 March 2001: http://www.psych.mcgill.ca/perpg/fac/shultz/default.htm Phone: 514 398-6139 Fax: 514 398-4896 -------------------------------------------------------- From vdavidsanchez at earthlink.net Sun Mar 25 19:22:55 2001 From: vdavidsanchez at earthlink.net (V. David Sanchez A.) Date: Sun, 25 Mar 2001 16:22:55 -0800 Subject: CFP NEUROCOMPUTING Special Issue on Neural Pattern Recognition Message-ID: <3ABE8BDF.3EAE761F@earthlink.net> CALL FOR PAPERS NEUROCOMPUTING An International Journal published by Elsevier Science B.V., vol. 36-41, 24 issues, in 2001 ISNN 0925-2312, URL: http://www.elsevier.nl/locate/neucom Special Issue on Neural Pattern Recognition Paper Submission Deadline: June 30th, 2001 Neural Pattern Recognition has been traditionally considered as an alternative approach to pattern recognition, different from the conventional statistical and structural approaches. There have been much success in practical pattern recognition applications using neural networks including Multilayer Perceptrons, Radial Basis Functions, and Self-Organizing Maps among others. Lately, there has been strong convergence between neural and statistical classifiers, exemplified by algorithms like Bayesian learning, the Expectation Maximization (EM) algorithm for mixture modeling, Support Vector Machines (SVM), or Independent Component Analysis (ICA) for feature extraction to mention just a few. The Neurocomputing journal invites original contributions for the forthcoming special issue on Neural Pattern Recognition from a broad scope of areas. Some topics relevant to this special issue include, but are not restricted to: -- Theoretical foundations -- Neural architectures for pattern recognition -- Hybrid recognition systems -- Relationship between neural and statistical classifiers -- Unsupervised learning in feature extraction -- Use of spiking neurons in pattern recognition -- Key applications including classification, clustering, data mining, bioinformatics, speech recognition, character recognition, 3-D object recognition, etc. Please send two hardcopies of the manuscript before June 30th, 2001, to: V. David Sanchez A., Neurocomputing - Editor in Chief - Advanced Computational Intelligent Systems P.O. Box 60130, Pasadena, CA 91116-6130 U.S.A. Street address: 1149 Wotkyns Drive Pasadena, CA 91103 Fax: +1-626-793-5120 Email: vdavidsanchez at earthlink.net including abstract, keywords, a cover page containing the title and author names, corresponding author name's complete address including telephone, fax, and email address, and clear indication to be a submission to the Special Issue on Neural Pattern Recognition. Guest Editors Krzysztof J. Cios University of Colorado at Denver Computer Science and Engineering Department U.S.A. Phone: (303) 556-4314 Fax: (303)556-8369 Email: kcios at carbon.cudenver.edu Kunihiko Fukushima The University of Electro-Communications Dept. of Information and Communication Engineering 1-5-1, Chofugaoka, Chyofu-shi, Tokyo 182-8585 Japan Fax +81-424-43-5291 Email: fukushima at ice.uec.ac.jp Erkki Oja, Helsinki University of Technology. Laboratory of Computer and Information Science P.O. Box 5400, 02015 HUT Finland Phone +358-9-4513265 Fax. +358-9-4513277 Email Erkki.Oja at hut.fi V. David Sanchez A., Neurocomputing - Editor in Chief - Advanced Computational Intelligent Systems P.O. Box 60130 Pasadena, CA 91116-6130 U.S.A. Fax: +1-626-793-5120 Email: vdavidsanchez at earthlink.net From steve at cns.bu.edu Tue Mar 27 11:26:18 2001 From: steve at cns.bu.edu (Stephen Grossberg) Date: Tue, 27 Mar 2001 08:26:18 -0800 Subject: frequency-dependent synaptic depression, Hebbian pairing, and synchrony in cortical pyramidal neurons Message-ID: The following article is now available at http://www.cns.bu.edu/Profiles/Grossberg in HTML, PDF, and Gzipped Postscript. Okatan, M. and Grossberg, S. Frequency-Dependent Synaptic Potentiation, Depression, and Spike Timing Induced by Hebbian Pairing in Cortical Pyramidal Neurons. Neural Networks ABSTRACT: Experiments by Markram and Tsodyks (1996) have suggested that Hebbian pairing in cortical pyramidal neurons potentiates or depresses the transmission of a subsequent pre-synaptic spike train at steady-state depending on whether the spike train is of low frequency or high frequency, respectively. The frequency above which pairing induced a significant decrease in steady-state synaptic efficacy was as low as about 20 Hz and this value depends on such synaptic properties as probability of release and time constant of recovery from short-term synaptic depression. These characteristics of cortical synapses have not yet been fully explained by neural models, notably the decreased steady-state synaptic efficacy at high pre-synaptic firing rates. This article suggests that this decrease in synaptic efficacy in cortical synapses was not observed at steady-state, but rather during a transition period preceding it whose duration is frequency-dependent. It is shown that the time taken to reach steady-state may be frequency-dependent, and may take considerably longer to occur at high than low frequencies. As a result, the pairing-induced decrease in synaptic efficacy at high pre-synaptic firing rates helps to localize the firing of the post-synaptic neuron to a short time interval following the onset of high frequency pre-synaptic spike trains. This effect may "speed up the time scale" in response to high frequency bursts of spikes, and may contribute to rapid synchronization of spike firing across cortical cells that are bound together by associatively learned connections. Key Words: synaptic potentiation, synaptic depression, frequency-dependent synaptic plasticity, cortical pyramidal cells, Hebbian pairing, cortical synchronization From jose at psychology.rutgers.edu Tue Mar 27 11:14:04 2001 From: jose at psychology.rutgers.edu (Stephen J. Hanson) Date: Tue, 27 Mar 2001 11:14:04 -0500 Subject: UNIX/LINUX GURUs! with Neural Net or Cognitive Neuroscience Interests.. Message-ID: <3AC0BC4C.1030803@kreizler.rutgers.edu> Departmental Systems Adminstration/UNIX/LINUX IMMEDIATE OPENING-- 04/01/01 PSYCHOLOGY DEPARTMENT-RUTGERS UNIVERSITY--Newark Campus--We are searching for an individual who can adminster the computing resources of the Psychology Department at Rutgers University (Newark). Resources include a network of SUN (Enterprise Server) /LINUX workstations, PCs and Macs, printers, voice mail system and various devices (scanners, projecters etc..). The individual will be responsible for installing and debugging software, and various routine system administration activites. Some proportion of time will be spent in research involving Cognitive Science especially related to Connectionist networks (or Neural Networks and Cognitive Neuroscience. Familiarity with C programming, UNIX system internals (BSD, System V, Solaris, Linux) and Windows (XX, NT) and local area networks running TCP/IP is required. Image processing or graphics programing experience are pluses. Candidates should possess either a BS/MS in Computer Science, Cognitive Science, AI or other relevant fields or equivalent experience. We are located 15-20 minutes from Downtown Manhattan and are within minutes of the NJPAC in the heart of Newark. Salary is competitive and will be dependent upon qualifications and experience. Rutgers University is an equal opportunity affirmative action employer. Please send snail mail resumes and references to Stephen J. Hanson Department of Psychology 101 Warren Street Rutgers University Newark, New Jersey, 07102 (Preferred) Direct email inquiries or resumes to: jose at psychology.rutgers.edu Please indicate on SUBJECT Line: SYS ADM as a keyword. From watrous at scr.siemens.com Tue Mar 27 13:07:11 2001 From: watrous at scr.siemens.com (Raymond L Watrous) Date: Tue, 27 Mar 2001 13:07:11 -0500 (EST) Subject: Career opportunity in signal processing R&D Message-ID: <200103271807.NAA03975@mt0.scr.siemens.com> CAREER OPPORTUNITY Senior Engineer Zargis Medical Corp. Princeton, NJ Zargis Medical Corp. has an immediate opening for a senior signal processing engineer. About Zargis Medical Zargis Medical Corp. is a newly formed company that develops and markets advanced diagnostic decision support products and services for primary care physicians and other healthcare professionals. Zargis Medical was formed through a joint investment by Siemens Corporate Research, a subsidiary of Siemens Corporation, and SPEEDUS.COM, Inc. (Nasdaq: SPDE). Job Description The senior engineer will provide leadership in developing advanced signal processing algorithms for the analysis and interpretation of heart sounds. The senior engineer will be involved in clinical data acquisition, product design, software development, quality assurance, and clinical trials. Qualifications The candidate should have an advanced degree in biomedical or electrical engineering, and a minimum of 5 years of experience in research and development. The candidate should possess superior mathematical ability and strong statistical signal processing skills. Familiarity with time-frequency methods, adaptive filtering, feature extraction, pattern matching, nonlinear optimization and hidden Markov models are essential. A background in biomedical signal processing or speech recognition would be an asset. The candidate should have demonstrated effectiveness in algorithm development, algorithm analysis and validation, technical communications and project management. Compensation Zargis Medical offers a competitive compensation and benefits plan. Zargis provides a congenial work environment and an outstanding career opportunity in an emerging technology company with excellent growth potential. Next Step Please forward a copy of your resume to: Raymond Watrous, Ph.D. Chief Technology Officer Zargis Medical Corp. 755 College Road East Princeton, NJ 08540 TEL: (609) 734-6596 FAX: (609) 734-6565 Email: careers at zargis.com From ismip00 at ee.usyd.edu.au Sat Mar 31 20:17:07 2001 From: ismip00 at ee.usyd.edu.au (ismip) Date: Sun, 1 Apr 2001 11:17:07 +1000 Subject: TNN Special Issue Final Call for Papers Message-ID: <004901c0ba49$7b794d20$471b4e81@pc.ee.usyd.edu.au> We apologize if you receive the message multiple times. Final Call for Papers IEEE Transaction on Neural Networks Special Issue on Intelligent Multimedia Processing Human communication is intrinsically multimodal. With the advances of technology, modern communication systems will also become more and more multimodal. Hence, multimedia technologies represent new ground for research interactions among a variety of media such as speech, audio, image, video, text and graphics. Future multimedia technologies will need to handle information with an increasing level of intelligence, i.e., automatic recognition and interpretation of multimodal signals. This is particularly emphasized in MPEG-7 which focuses on the 'multimedia content description interface'. The description shall be associated with the content itself to facilitate fast and effective searching for all the media. Specifically, the MPEG-7 research domain will cover techniques for content-based indexing and retrieval: pattern recognition, face detection/recognition, and fusion of multimodality. Intelligent multimedia processing shares three fundamental goals with biological systems: a) Universal data processing engine for multimodal signals; b) Multimodality; and c) Unsupervised clustering and/or supervised learning by examples. Because of these features, neural networks are attractive candidates for intelligent multimedia processing and recent activity in the area is a proof of this fact. The main attribute of neural computing is its adaptive learning capability, which enables interpretations of possible variations of a same object or pattern, e.g., with respect to scale, orientation, and perspective. Moreover, they are able to accurately approximate unknown systems based on sparse sets of noisy data. Certain neural models also effectively incorporate statistical signal processing and optimization techniques. In addition, spatial/temporal neural structures and hierarchical models are promising for multirate, multiresolution multimedia processing. As a result, many successful applications of neural networks in intelligent multimedia processing, sometimes combined with fuzzy systems and evolutionary computing, have been reported. The possible topics for the special issue include, but are not limited to, the following: * Neural networks (including BSS and ICA) and other computational intelligence models, learning paradigms, and architectures for multimedia processing. * Intelligent multimedia processing architectures. * Multimedia/multichannel data fusion. * Multimodal representation and information retrieval: Applications in hyperlinking of multimedia objects, query and search of multimedia information including intelligent web agents, 3D object representation and motion tracking, image sequence generation and animation. * Human-computer interaction and communications: face recognition, lip-reading analysis, facial expression and emotion categorization, interactive human-machine vision, speech recognition, speaker recognition, gesture analysis and recognition, auditory/visual scene analysis, and multimodal interaction. * Multimedia data analysis and visualization: texture, color, content, etc. * Intelligent network control of audio/video streams in multimedia networking applications. Original, previously-unpublished research articles as well as state-of-the-art tutorial papers will be considered. Authors should follow the IEEE TNN manuscript format described in the Information for Authors, which can be found on the inside back cover of any issue of TNN. Prospective authors are invited to submit papers to the website: http://eivind.imm.dtu.dk/tnn. The following schedule will apply: Manuscript submission: April 15, 2001 Acceptance notification: July 31, 2001 Final manuscripts due: October 30, 2001 Publication: January 2002 Guest Editors: Tulay Adali, Ling Guan, Dept of CSEE School of Electrical & Information Eng. Univ of Maryland, Baltimore County The University of Sydney Baltimore, MD 21250 Sydney, NSW 2000 Australia Jan Larsen Shigeru Katagiri Dept of Mathematical Modelling ATR Technical University of Denmark 2-2 Hikaridai 2800 Lyngby Seika-cho, Soraku-gun Denmark Kyoto 619-02 Japan Jose Principe Dept of Electrical & Computer Eng University of Florida Gainesville, FL 32611 From m.niranjan at dcs.shef.ac.uk Thu Mar 1 11:51:56 2001 From: m.niranjan at dcs.shef.ac.uk (Mahesan Niranjan) Date: Thu, 1 Mar 2001 16:51:56 +0000 (GMT) Subject: Workshop on Geometric Computations In-Reply-To: <200103011100.LAA13013@padley.dcs.shef.ac.uk> Message-ID: +++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++ UNCERTAINTY IN GEOMETRIC COMPUTATIONS, 5-6 July 2001, Sheffield, England Invited Speakers (will be expanded) : Shun-ichi Amari (RIKEN, Japan), Andrew Blake (Microsoft, UK), Adrian Bowyer (Bath, UK), Alan Edelman (MIT, USA), Robin Forrest (East Anglia, UK), Nicholas Higham (Manchester, UK), Dinesh Manocha (North Carolina, USA), Si Wu (Sheffield, UK) Organisers: Joab Winkler and Mahesan Niranjan Department of Computer Science The University of Sheffield, UK. The representation and management of uncertainty is an important issue in several different disciplines, such as numerical problems in computer graphics that occur when calculating the intersection curve of two surfaces, high performance pattern classification in a feature space, and the study of families of probability distributions in information geometry. The aim of this two-day workshop is to explore the underlying geometric theme that is common to these diverse disciplines. The workshop will consist of a number of invited contributions of a tutorial nature covering the different topics, contributed papers from participants and discussion sessions that explore the connections. Contributions will be published by Kluwer in an edited volume. The workshop is sponsored by the EPSRC and LMS, and financial support is available to cover costs of UK based graduate students. The total number of participants is limited to 70. One page abstracts are invited from potential participants. Please submit electronically (postscript, PDF or plain text) to Dr Joab Winkler Deadline for Abstracts: 15 April 2001 For further information see: http://www.shef.ac.uk/~geom2001/ +++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++ From wolfskil at MIT.EDU Thu Mar 1 15:51:02 2001 From: wolfskil at MIT.EDU (Jud Wolfskill) Date: Thu, 1 Mar 2001 15:51:02 -0500 Subject: book announcement--Spirtes Message-ID: I thought readers of the Connectionist List might be interested in this book. For more information please visit http://mitpress.mit.edu/promotions/books/SPICHF00. Thank you! Best, Jud Causation, Prediction, and Search second edition Peter Spirtes, Clark Glymour, and Richard Scheines What assumptions and methods allow us to turn observations into causal knowledge, and how can even incomplete causal knowledge be used in planning and prediction to influence and control our environment? In this book Peter Spirtes, Clark Glymour, and Richard Scheines address these questions using the formalism of Bayes networks, with results that have been applied in diverse areas of research in the social, behavioral, and physical sciences. The authors show that although experimental and observational study designs may not always permit the same inferences, they are subject to uniform principles. They axiomatize the connection between causal structure and probabilistic independence, explore several varieties of causal indistinguishability, formulate a theory of manipulation, and develop asymptotically reliable procedures for searching over equivalence classes of causal models, including models of categorical data and structural equation models with and without latent variables. The authors show that the relationship between causality and probability can also help to clarify such diverse topics in statistics as the comparative power of experimentation versus observation, Simpson's paradox, errors in regression models, retrospective versus prospective sampling, and variable selection. The second edition contains a new introduction and an extensive survey of advances and applications that have appeared since the first edition was published in 1993. Peter Spirtes is Professor of Philosophy at the Center for Automated Learning and Discovery, Carnegie Mellon University. Clark Glymour is Alumni University Professor of Philosophy at Carnegie Mellon University and Valtz Family Professor of Philosophy at the University of California, San Diego. He is also Distinguished External Member of the Center for Human and Machine Cognition at the University of West Florida, and Adjunct Professor of Philosophy of History and Philosophy of Science at the University of Pittsburgh. Richard Scheines is Associate Professor of Philosophy at the Center for Automated Learning and Discovery, and at the Human Computer Interaction Institute, Carnegie Mellon University. 7 x 9, 496 pp., 225 illus., cloth ISBN 0-262-19440-6 Adaptive Computation and Machine Learning series A Bradford Book -------------------------------------------------------------------------------- Jud Wolfskill 617.253.2079 phone Associate Publicist 617.253.1709 fax MIT Press wolfskil at mit.edu 5 Cambridge Center http://mitpress.mit.edu Fourth Floor Cambridge, MA 02142 From harnad at coglit.ecs.soton.ac.uk Fri Mar 2 11:15:43 2001 From: harnad at coglit.ecs.soton.ac.uk (Stevan Harnad) Date: Fri, 2 Mar 2001 16:15:43 +0000 (GMT) Subject: Review of: Damasio, Edelman, McGinn, Tomasello and Fodor Message-ID: Review of: Antonio Damasio: The Feeling of What Happens Gerald M. Edelman and Giulio Tononi: A Universe of Consciousness Colin McGinn The Mysterious Flame Michael Tomasello Cultural Origins of Human Cognition Jerry A. Fodor The Mind Doesn't Work That Way To appear in The Sciences New York Academy of Sciences, April 2001 k http://www.cogsci.soton.ac.uk/~harnad/Tp/bookrev.htm -------------------------------------------------------------------- Stevan Harnad harnad at cogsci.soton.ac.uk Professor of Cognitive Science harnad at princeton.edu Department of Electronics and phone: +44 23-80 592-582 Computer Science fax: +44 23-80 592-865 University of Southampton http://www.cogsci.soton.ac.uk/~harnad/ Highfield, Southampton http://www.princeton.edu/~harnad/ SO17 1BJ UNITED KINGDOM From ted.carnevale at yale.edu Sun Mar 4 19:27:01 2001 From: ted.carnevale at yale.edu (Ted Carnevale) Date: Sun, 04 Mar 2001 19:27:01 -0500 Subject: NEURON 2001 Summer Course Message-ID: <3AA2DD55.8343479E@yale.edu> COURSE ANNOUNCEMENT What: "The NEURON Simulation Environment" (NEURON 2001 Summer Course) When: Saturday, June 23, through Wednesday, June 27, 2001 Where: San Diego Supercomputer Center University of California at San Diego, CA Organizers: N.T. Carnevale and M.L. Hines Faculty includes: N.T. Carnevale, M.L. Hines, W.W. Lytton, and T.J. Sejnowski Description: This intensive hands-on course covers the design, construction, and use of models in the NEURON simulation environment. It is intended primarily for those who are concerned with models of biological neurons and neural networks that are closely linked to empirical observations, e.g. experimentalists who wish to incorporate modeling in their research plans, and theoreticians who are interested in the principles of biological computation. The course is designed to be useful and informative for registrants at all levels of experience, from those who are just beginning to those who are already quite familiar with NEURON or other simulation tools. Registration is limited to 20, and the deadline for receipt of applications is Friday, May 25, 2001. For more information see http://www.neuron.yale.edu/sdsc2001/sdsc2001.htm or contact Ted Carnevale Psychology Dept. Box 208205 Yale University New Haven, CT 06520-8205 USA phone 203-432-7363 fax 203-432-7172 email ted.carnevale at yale.edu Supported in part by: National Science Foundation National Institutes of Health National Partnership for Advanced Computational Infrastructure and the San Diego Supercomputer Center Contractual terms require inclusion of the following statement: This course is not sponsored by the University of California. --Ted From juergen at idsia.ch Mon Mar 5 12:24:11 2001 From: juergen at idsia.ch (juergen@idsia.ch) Date: Mon, 5 Mar 2001 18:24:11 +0100 Subject: optimal search algorithm Message-ID: <200103051724.SAA27997@ruebe.idsia.ch> Dear connectionists, Recently Marcus Hutter has developed a very general, asymptotically optimal search method that should be of interest to researchers in the areas of reinforcement learning & neural networks & AI. The method is not limited to machine learning problems though - I believe it will find its way into general computer science textbooks. With the benefit of hindsight I find it amazing that it has remained undiscovered up until the turn of the century. ------------------------------------------------------------------------- The fastest and shortest algorithm for all well-defined problems Marcus Hutter, IDSIA An algorithm M is described that solves any well-defined problem p as quickly as the fastest algorithm computing a solution to p, save for a factor of 5 and low-order additive terms. M optimally distributes resources between the execution of provably correct p-solving programs and an enumeration of all proofs, including relevant proofs of program correctness and of time bounds on program runtimes. M avoids Blum's speed-up theorem by ignoring programs without correctness proof. M has broader applicability and can be faster than Levin's universal search, the fastest method for inverting functions save for a large multiplicative constant. An extension of Kolmogorov complexity and two novel natural measures of function complexity are used to show that the most efficient program computing some function f is also among the shortest programs provably computing f. ftp://ftp.idsia.ch/pub/techrep/IDSIA-16-00.ps.gz ------------------------------------------------------------------------- Juergen Schmidhuber www.idsia.ch From t.c.pearce at leicester.ac.uk Mon Mar 5 13:20:56 2001 From: t.c.pearce at leicester.ac.uk (Tim Pearce) Date: Mon, 5 Mar 2001 18:20:56 -0000 Subject: Postdoctoral Research Positions Message-ID: Three Postdoctoral Research Positions Silicon Olfactory System Implementation We seek applications for three postdoctoral RAs to form the core team of a large EPSRC funded project to develop the worlds first silicon implementation of the olfactory system. Work will be aimed towards producing a micropower, fully integrated, electronic nose, with the capability of adapting to changes in its operating conditions. This is a major project conducted between three UK Universities and, Osmetech PLC a leading international chemical sensing instrumentation entity. There will be various opportunities for presenting work at international conferences, as well as opportunities for travel to our Swiss collaborators at ETH, Zrich. More details on the project are available at http://www.le.ac.uk/eg/tcp1/avlsi/ 1. Computational Neuroscientist will be responsible for developing a detailed computational model of the olfactory bulb, constrained by biological data. The model will be driven by novel ChemFET array technology developed at Warwick and finally will be implemented in aVLSI at Edinburgh. The applicant will ideally have modeling experience in neuronal systems (but not necessarily the olfactory system), a PhD in a related area and strong mathematical skills. For more information on the overall project and this position in particular contact Dr. Tim Pearce, Department of Engineering, University of Leicester, University Road, Leicester LE1 7RH, UK, +44 (0) 116 223 1290, t.c.pearce at le.ac.uk 2. Silicon Sensor/MEMS Specialist will be responsible for the design and fabrication of a novel odour sensing structure based upon microfluidic channels carrying the odour to an array of polymer-coated MOSFET devices. The applicant will have a PhD in electronics and ideally some experience in using device layout and simulation software, such as Tanner Tools or Cadence, Medici, ISE. For more information on the sensors position contact Prof. Julian Gardner, School of Engineering, University of Warwick, Coventry CV4 7AL, UK +44 (0) 24 76523 695, j.w.gardner at warwick.ac.uk 3. Analogue VLSI Engineer. S/He will be responsible for designing the on chip analogue interface to the ChemFET sensor array technology developed at Warwick and implementing the computational model of the olfactory bulb developed at Leicester. The applicant will ideally have some experience in analogue circuit design, good written and verbal communication skills and a PhD in a related subject area. For more details contact Dr. Alister Hamilton, Department of Electronics and Electrical Engineering, University of Edinburgh, Mayfield Road, Edinburgh EH9 3JL, UK, +44 (0) 131 650 5597, Alister.Hamilton at ee.ed.ac.uk Salary range will be on the RA1A scale (16,775 - 25,213 depending upon age and experience). The RAs will be employed for up to three years. Applications should be addressed to the respective academic, including a CV, list of publications and the name of two referees, who should be asked to send their reference to the same address). Applications may also be e-mailed. Closing date: 1st July, 2001. -- T.C. Pearce, PhD URL: http://www.leicester.ac.uk/eg/tcp1/ Lecturer in Bioengineering E-mail: t.c.pearce at leicester.ac.uk Department of Engineering Tel: +44 (0)116 223 1290 University of Leicester Fax: +44 (0)116 252 2619 Leicester LE1 7RH Bioengineering, Transducers and United Kingdom Signal Processing Group From ingber at ingber.com Mon Mar 5 20:30:55 2001 From: ingber at ingber.com (Lester Ingber) Date: Mon, 5 Mar 2001 19:30:55 -0600 Subject: Paper: Probability tree algorithm for general diffusion processes Message-ID: <20010305193055.A17296@ingber.com> The following preprint is available: %A L. Ingber %A C. Chen %A R.P. Mondescu %A D. Muzzall %A M. Renedo %T Probability tree algorithm for general diffusion processes %D 2001 %O URL http://www.ingber.com/path01_pathtree.ps.gz ABSTRACT Motivated by path-integral numerical solutions of diffusion processes, PATHINT, we present a new tree algorithm, PATHTREE, which permits extremely fast accurate computation of probability distributions of a large class of general nonlinear diffusion processes. -- Prof. Lester Ingber http://www.ingber.com/ PO Box 06440 Sears Tower Chicago IL 60606-0440 http://www.alumni.caltech.edu/~ingber/ From terry at salk.edu Wed Mar 7 17:48:31 2001 From: terry at salk.edu (Terry Sejnowski) Date: Wed, 7 Mar 2001 14:48:31 -0800 (PST) Subject: NEURAL COMPUTATION 13:4 Message-ID: <200103072248.f27MmVH58010@kepler.salk.edu> Neural Computation - Contents - Volume 13, Number 4 - April 1, 2001 ARTICLE Sampling Properties of the Spectrum and Coherency of Sequences of Action Potentials M. R. Jarvis and P. P. Mitra NOTE A Novel Spike Distance M. C. W. van Rossum LETTERS On Synchrony of Weakly Coupled Neurons at Low Firing Rates L. Neltner and David Hansel Population Coding with Correlation and an Unfaithful Model Si Wu, Hiroyuki Nakahara and Shun-ichi Amari Metabolically Efficient Information Processing Vijay Balasubramanian, Don Kimber and Michael J. Berry II Effective Neuronal Learning with Ineffective Hebbian Learning Rules Gal Chechik, Isaac Meilijson, and Eytan Ruppin Internal Model Reproduces Anticipatory Neural Activity Roland E. Suri and Wolfram Schultz Blind Source Separation by Sparse Decomposition Michael Zibulevsky and Barak A. Pearlmutter Complexity Pursuit: Separating Interesting Components from Time-Series Aapo Hyvarinen Algebraic Analysis for Non-Identifiable Learning Machines Sumio Watanabe A New On-Line Learning Model Shahar Mendelson ----- ON-LINE - http://neco.mitpress.org/ SUBSCRIPTIONS - 2001 - VOLUME 13 - 12 ISSUES USA Canada* Other Countries Student/Retired $60 $64.20 $108 Individual $88 $94.16 $136 Institution $460 $492.20 $508 * includes 7% GST MIT Press Journals, 5 Cambridge Center, Cambridge, MA 02142-9902. Tel: (617) 253-2889 FAX: (617) 577-1545 journals-orders at mit.edu ----- From doya at isd.atr.co.jp Thu Mar 8 04:21:24 2001 From: doya at isd.atr.co.jp (Kenji Doya) Date: Thu, 8 Mar 2001 18:21:24 +0900 Subject: CREST Workshop on Metalearning and Neuromodulation Message-ID: Dear connectionists, We are organizing the following workshop. The program and the abstracts are posted on the web page. If you are interested, please register as soon as possible. We still have a few openings for poster presentations. Best wishes, Kenji Doya ************************************************************************ CREST WORKSHOP ON METALEARNING AND NEUROMODULATION April 6th and 7th, 2001 Keihanna Plaza, Seika, Kyoto, Japan Sponsored by Metalearning, Neuromodulation, and Emotion Project CREST, Japan Science and Technology Corporation The goal of this workshop is to bring together neuroscientists as well as theorists who work on the regulatory mechanisms of adaptive systems, either biological or artificial. Neuromodulators such as dopamine, serotonin, norepinephrine, and acetylcholine have widespread influences on information processing in the brain. Recent neurobiological studies revealed their specific roles such as prediction of future reward and punishment, regulation of accuracy and diversity of behaviors, and the rate of memory update. Computational studies of 'metalearning,' the way of adapting global parameters and structures in a learning system, can be helpful in understanding the roles and the dynamics of neuromodulators. We invite researchers who are working on the the biological mechanisms of neuromodulators and the computational theory of metalearning, as well as those who are interested in the neurochemical basis and computational mechanisms of the mind. Please visit our home page for details: http://www.isd.atr.co.jp/nip/crest/workshop/ This workshop is scheduled after the 9th International Catecholamine Symposium (http://mc-net.jtbcom.co.jp/ics2001/) held in downtown Kyoto from March 31st to April 5th, 2001. INVITED SPEAKERS Minoru Asada, Osaka University Gary Aston-Jones, University of Pennsylvania Kenji Doya, CREST, JST & ATR International Barry J. Everitt, University of Cambridge Michael Hasselmo, Boston University Okihide Hikosaka, Juntendo University Tadashi Isa, National Institute for Physiological Sciences Shin Ishii, Nara Institute of Science and Technology Sham Kakade, University College London Takashi Matsumoto, Waseda University Toshiyuki Sawaguchi, Hokkaido University Wolfram Schultz, University of Fribourg Yuko Sekino, Gumma University Shigeto Yamawaki, Hiroshima University CALL FOR POSTERS A poster session will be held in the evening of April 6th. Please send 1) title, 2) author(s), 3) affiliation(s), 4) abstract up to 500 words, 5) postal address, 6) phone, 7) fax, and 8) e-mail address of the presenting author by e-mail to the secretariat. REGISTRATION Please send 1) name, 2) affiliation, 3) postal address, 4) phone, 5) fax, and 6) e-mail address by e-mail to the secretariat by March 15, 2001. Registration is free. However, the numbers of attendees will be limited by capacity of the conference room. So please register early. HOTEL RESERVATION Please visit our home page (http://www.isd.atr.co.jp/nip/crest/workshop/) to download a reservation form, and send it by fax directly to a hotel. Note that early April is the high season for cherry blossoms in Kyoto, so please reserve early. SECRETARIAT Naomi Katayama Metalearning, Neuromodulation and Emotion, CREST, JST 2-2-2 Hikaridai, Seika, Soraku, Kyoto 619-0288, Japan Phone: +81-774-95-1251 Fax: +81-774-95-1259 E-mail: nip-info at isd.atr.co.jp URL: http://www.isd.atr.co.jp/nip/crest/ ---- Kenji Doya Information Sciences Division, ATR International; CREST, JST 2-2-2 Hikaridai, Seika, Soraku, Kyoto 619-0288, Japan Phone:+81-774-95-1251; Fax:+81-774-95-1259; http://www.isd.atr.co.jp/~doya From rsun at pc113.cecs.missouri.edu Fri Mar 9 19:09:36 2001 From: rsun at pc113.cecs.missouri.edu (Dr. Ron Sun) Date: Fri, 9 Mar 2001 18:09:36 -0600 Subject: new papers on cognitive models available Message-ID: <200103100009.f2A09aG12197@pc113.cecs.missouri.edu> Announcing several papers on cognitive modeling and cognitive architectures based on hybrid reinforcement learning --- the CLARION model: A paper on cognitive modeling using CLARION: -------------------------------------------------- From sgg at copper.dcs.qmw.ac.uk Mon Mar 12 07:35:57 2001 From: sgg at copper.dcs.qmw.ac.uk (sgg@copper.dcs.qmw.ac.uk) Date: Mon, 12 Mar 2001 12:35:57 +0000 (GMT) Subject: Post-doc positions in computer vision Message-ID: <200103121235.MAA26547@dcs.qmw.ac.uk> COMPUTER VISION RESEARCH: UNDERSTANDING VISUAL BEHAVIOUR DEPARTMENT OF COMPUTER SCIENCE QUEEN MARY, UNIVERSITY OF LONDON U.K. Postdoctoral Research Assistants (2 posts) Applications are invited for two Post-Doctoral Research Assistants to work on a computer vision project funded by EPSRC, DTI and industry on the recognition of visual behaviour in video data. In addition to Queen Mary, University of London, the project consortium consists of a number of industrial partners including Safehouse Technologies, BAA (former British Airport Authorities), BT Labs, and Heritage Protection. You will undertake novel research in dynamic scene understanding, zero-motion recognition, visual events modelling, behaviour profiling and abnormal behaviour recognition. You should have experience in computer vision and statistical learning research. Any experience in processing image sequence data from live video cameras would be an advantage. You should also be competent in programming with C or C++ within X and NT environments. More details about computer vision research at Queen Mary can be found at http://www.dcs.qmw.ac.uk/research/vision/. Both posts are for 2 years and are to start as soon as possible. Salary are in range of 20,865 pounds - 22,599 pounds per annum inclusive, depending on experience. For further details and an application form, please phone the Department of Computer Science, Queen Mary, University of London on 020-7882 5227 quoting reference number 01046, or contact Shaogang Gong at sgg at dcs.qmw.ac.uk or Dennis Parkinson at dennisp at dcs.qmw.ac.uk. Completed application forms should be returned by Wednesday 4th April 2001 to Ms. Gill Carter, Department of Computer Science, Queen Mary, University of London, London, E1 4NS. QUEEN MARY: WORKING TOWARDS EQUAL OPPORTUNITIES From aco-list at iridia0.ulb.ac.be Mon Mar 12 11:34:24 2001 From: aco-list at iridia0.ulb.ac.be (aco-list@iridia0.ulb.ac.be) Date: Mon, 12 Mar 2001 17:34:24 +0100 (CET) Subject: Ant Colony Optimization mailing list Message-ID: <200103121634.RAA21569@iridia0.ulb.ac.be> Dear Colleagues, we would like to announce that a moderated digest about Ant Colony Optimization and Ant Algorithms has been set up. It collects news, information, new ideas and comments about the above-mentioned research area. Main topics: - Ant colony optimization - Ant algorithms - Swarm intelligence and insect behavior - Bio-inspired multi-agent systems Relevant information for the ACO-list are: - Conferences and workshops - Recent pubblications - New products (not just for commercial purposes, but with scientific interest) - Books and articles reviews - Open positions and any other information which could be useful for researchers in this area. For SUBSCRIPTIONS write to: ACO-list-request at iridia.ulb.ac.be For SUBMISSIONS write to: ACO-list at iridia.ulb.ac.be Marco Dorigo and Andrea Roli (list moderators) From Martin.Appl at mchp.siemens.de Mon Mar 12 06:46:12 2001 From: Martin.Appl at mchp.siemens.de (Martin Appl) Date: Mon, 12 Mar 2001 12:46:12 +0100 Subject: Ph.D. thesis available: Model-Based Reinforcement Learning in Continuous Environments Message-ID: <000101c0aaea$0a8dd3b0$0dba178b@mhpaaxsc.mchp.siemens.de> Dear Connectionists, my Ph.D. thesis **************************************************************** Model-Based Reinforcement Learning in Continuous Environments Martin Appl December 2000, Technical University of Munich **************************************************************** is now available at www.martinappl.de . Best regards, Martin Appl ++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++ ABSTRACT Reinforcement learning enables machines to learn from experiences. For example, controllers can learn optimal control strategies trying out different strategies and evaluating the resulting performance of the processes under control. At present reinforcement learning is rarely used for the optimization of complex industrial processes, since the computational requirements of reinforcement learning approaches grow fast as the number of input variables increases. Hence, the rough goal of this thesis is to develop efficient approaches enhancing the range of application of reinforcement learning. The focus of the thesis is on time-discrete control of processes with continuous controlled inputs and continuous measured outputs. A central result of this thesis is a fuzzy model-based reinforcement learning approach. Using this approach control strategies for continuous processes can be efficiently trained. The output of the approach is a Takagi-Sugeno fuzzy system representing the optimal control strategy. A further result of this thesis is a fuzzy model-based exploration strategy. During learning this strategy controls processes in such a way that maximum information is gained. Hence, the number of control cycles required to learn optimal control strategies is significantly reduced. For many control problems it is known a priori which measured quantities are correlated and which are statistically independent. Taking this kind of a priori knowledge into account both the model-based learning approach and the exploration strategy can be significantly sped up, as is also shown in this thesis. A general problem in fuzzy model-based learning is the generation of suitable fuzzy partitions. Defining partitions by hand is not trivial, since fine partitions lead to a large number of states, whereas coarse partitions can be unsuitable for the representation of the optimal control strategy. Therefore, further extensions of the fuzzy model-based learning approach and the model-based exploration strategy are presented in this thesis. The basic idea behind these extensions is to represent internal models by clustered transitions. Based on this compact representation the extended algorithms can automatically determine suitable partitions of the state space. The methods presented in this thesis are applied to tasks from traffic signal control. One task is to select framework signal plans in dependence of traffic conditions. It turned out that the fuzzy model-based approaches outperform existing crisp methods. Furthermore, these methods allow to solve the task in reasonable time. From mieko at isd.atr.co.jp Tue Mar 13 05:06:59 2001 From: mieko at isd.atr.co.jp (Mieko Namba) Date: Tue, 13 Mar 2001 19:06:59 +0900 Subject: Neural Networks 14(3) Message-ID: NEURAL NETWORKS 14(3) Contents - Volume 14, Number 3 - 2001 ------------------------------------------------------------------ LETTERS: New theorems on global convergence of some dynamical systems. T. Chen, S. Amari INVITED ARTICLE: Bayesian approach for neural networks - review and case studies. J. Lampinen, A. Vehtari CONTRIBUTED ARTICLES: ***** Neuroscience and Neuropsychology ***** Temporal clustering with spiking neurons and dynamic synapses: towards technological applications. J. Storck, F. Jakel, G. Deco ***** Mathematical and Computational Analysis ***** An evaluation of standard retrieval algorithms and a binary neural approach. V.J. Hodge, J. Austin Pseudo-outer product based fuzzy neural network fingerprint verification system. C. Quek, K. B. Tan, V. K. Sagar ***** Engineering and Design ***** A what-and-where fusion neural network for recognition and tracking of multiple radar emitters. E. Granger, M. A. Rubin, S. Grossberg, P. Lavoie ***** TECHNOLOGY & APPLICATIONS ***** Graph matching vs mutual information maximization for object detection. L.B. Shams, M. J. Brady, S. Schaal Naural networks for improved target differentiation and localization with sonar. B. Ayrulu, B. Barshan ------------------------------------------------------------------ Electronic access: www.elsevier.com/locate/neunet/. Individuals can look up instructions, aims & scope, see news, tables of contents, etc. Those who are at institutions which subscribe to Neural Networks get access to full article text as part of the institutional subscription. Sample copies can be requested for free and back issues can be ordered through the Elsevier customer support offices: nlinfo-f at elsevier.nl usinfo-f at elsevier.com or info at elsevier.co.jp ------------------------------ INNS/ENNS/JNNS Membership includes a subscription to Neural Networks: The International (INNS), European (ENNS), and Japanese (JNNS) Neural Network Societies are associations of scientists, engineers, students, and others seeking to learn about and advance the understanding of the modeling of behavioral and brain processes, and the application of neural modeling concepts to technological problems. Membership in any of the societies includes a subscription to Neural Networks, the official journal of the societies. Application forms should be sent to all the societies you want to apply to (for example, one as a member with subscription and the other one or two as a member without subscription). The JNNS does not accept credit cards or checks; to apply to the JNNS, send in the application form and wait for instructions about remitting payment. The ENNS accepts bank orders in Swedish Crowns (SEK) or credit cards. The INNS does not invoice for payment. ---------------------------------------------------------------------------- Membership Type INNS ENNS JNNS ---------------------------------------------------------------------------- membership with $80 or 660 SEK or Y 15,000 [including Neural Networks 2,000 entrance fee] or $55 (student) 460 SEK (student) Y 13,000 (student) [including 2,000 entrance fee] ----------------------------------------------------------------------------- membership without $30 200 SEK not available to Neural Networks non-students (subscribe through another society) Y 5,000 (student) [including 2,000 entrance fee] ----------------------------------------------------------------------------- Institutional rates $1132 2230 NLG Y 149,524 ----------------------------------------------------------------------------- Name: _____________________________________ Title: _____________________________________ Address: _____________________________________ _____________________________________ _____________________________________ Phone: _____________________________________ Fax: _____________________________________ Email: _____________________________________ Payment: [ ] Check or money order enclosed, payable to INNS or ENNS OR [ ] Charge my VISA or MasterCard card number ____________________________ expiration date ________________________ INNS Membership 19 Mantua Road Mount Royal NJ 08061 USA 856 423 0162 (phone) 856 423 3420 (fax) innshq at talley.com http://www.inns.org ENNS Membership University of Skovde P.O. Box 408 531 28 Skovde Sweden 46 500 44 83 37 (phone) 46 500 44 83 99 (fax) enns at ida.his.se http://www.his.se/ida/enns JNNS Membership c/o Professor Tsukada Faculty of Engineering Tamagawa University 6-1-1, Tamagawa Gakuen, Machida-city Tokyo 113-8656 Japan 81 42 739 8431 (phone) 81 42 739 8858 (fax) jnns at jnns.inf.eng.tamagawa.ac.jp http://jnns.inf.eng.tamagawa.ac.jp/home-j.html ----------------------------------------------------------------- From Dominique.Martinez at loria.fr Tue Mar 13 12:01:32 2001 From: Dominique.Martinez at loria.fr (Dominique Martinez) Date: Tue, 13 Mar 2001 18:01:32 +0100 (MET) Subject: Postdoctoral Fellowship in Neuromimetic Olfatory SEnsing Message-ID: Postdoctoral Fellowship in Neuromimetic Olfatory SEnsing (NOSE). A one year and a half postdoctoral fellowship is available immediately within the cooperative research project NOSE from INRIA. The first objective of this project will be to develop biologically inspired spiking neural network models of olfactory perception. The second objective will be to apply these models in an autonomous robot so as to mimic the animal behavior of tracking specific odours. The postdoctoral fellow will join the CORTEX group at LORIA-INRIA in Nancy, France. For further information see http://www.loria.fr/~rochel/nose or contact Dominique Martinez CORTEX group - LORIA Campus Scientifique - BP 239 54506 Vandoeuvre-Les-Nancy E-mail: dmartine at loria.fr Phone: (+33) 3-83-59-30-72 Fax: (+33) 3-83-41-30-79 From gomi at idea.brl.ntt.co.jp Tue Mar 13 20:36:54 2001 From: gomi at idea.brl.ntt.co.jp (Hiroaki GOMI) Date: Wed, 14 Mar 2001 10:36:54 +0900 Subject: RESEARCH POSITIONS AVAILABLE at NTT Communication Sci. Labs. Message-ID: <20010314103654Z.gomi@idea.brl.ntt.co.jp> RESEARCH POSITIONS AVAILABLE NTT (Nippon Telegraph and Telephone Corporation) Communication Science Labs. Research associate (post doctoral fellow) and research specialist positions are currently available in the human speech & motor control group at NTT Communication Science Laboratories. One research topic focuses on the human motor control mechanisms in the mechanical interaction with environments, and its brain information processing. The candidate should have a strong background in human motor control studies. Some skills of C programming for controlling experimental system and Matlab for data analyses are desirable but not required. Candidates with training in psychological approaches to motor control are particularly encouraged to apply. Another topic focuses on information processing for speech and language. The candidate should have a strong background in the fields of speech and language. The position requires strong programming skills in C and/or C++ with experience in programming mathematical algorithms, and knowledge of signal processing and machine learning. The salary will be decided according to company regulation. The research center for these topics is located in Atsugi near Tokyo, Japan (see our web page). The positions are tenable for one year and are renewable every year until the limit of two years. Please send curriculum vitae including research and programming experiences, names and contact details of referees, and representative publications to the address listed below. Curriculum vitae should be sent until March 23, 2001. Informal inquiries should be set to the e-mail address. E-mail inquiries and applications are encouraged. We prefer a starting date of April 1, 2001 but this is open to negotiation. Application should be sent to: Dr. Eisaku Maeda soukatsu at cslab.kecl.ntt.co.jp Senior Manager of NTT Communication Science Laboratories Tel: +81-774-93-5040 Fax: +81-774-93-5045 address: NTT Communication Science Labs. Hikaridai 2-4, Seika-cho, Soraku-gun, Kyoto-pref. 619-0237, Japan web page: http://www.kecl.ntt.co.jp From swatanab at pi.titech.ac.jp Wed Mar 14 20:44:29 2001 From: swatanab at pi.titech.ac.jp (Sumio Watanabe) Date: Thu, 15 Mar 2001 10:44:29 +0900 Subject: CFP: Special Session in KES'2001: Geometry and Statistics in NN Message-ID: <004f01c0acf1$79fd19a0$918a7083@deskpowerts> Dear Connectionists, We have a special session on Geometry and Statistics in NN in the international conference KES'2001, which will be held in Oska and Nara in Japan, 6,7,8, September, 2001. http://www.bton.ac.uk/kes/kes2001/ In the session "Geometry and Statistics in Neural Network Learning Theory", we study the statistical problem caused by non-identifiability of layered learning machines. (See the followings). We would like to invite some researchers who will take part in this session and present a paper. http://www.bton.ac.uk/kes/kes2001/sessions.html. The authors who will be invited: Shun-Ichi Amari (RIKEN Brain Science Institute,Japan) Kenji Fukumizu (Insitute of Statistical Mathematics,Japan) Katsuyuki Hagiwara (Mie University,Japan) Sumio Watanabe(Tokyo Institute of Technology,Japan) If you are interested in this special session, please contact the following e-mail address by March 31th, 2001. swatanab at pi.titech.ac.jp Thank you very much. Sincerely, Dr. Sumio Watanabe, Associate Professor Advanced Information Processing Division Precision and Intelligence Laboratory Tokyo Institute of Technology (Fax) +81-45-924-5018 E-mail: swatanab at pi.titech.ac.jp http://watanabe-www.pi.titech.ac.jp/~swatanab/index.html ***************Special Session *************** Geometry and Statistics in Neural Network Learning Theory Non-identifiability: A learning machine p(x|w) with a parameter w is called identifiable if the mapping from w to p(x|w) is one-to-one. It should be emphasized that almost all learning machines with hidden parts are not identifiable, resulting that the manifolds of parameters have singular Fisher metrics. Problem: If a non-identifiable learning machine can almost approximate the true distribution, then, because of the finiteness of training samples, it is often in an almost redundant state. In such a case, neither the distribution of the MLE nor the Bayes a posteriori distrbution is subject to the asymptotically normal distribution. The conventional statistical asymptotic theory can not be applied to analyze the learning curves. Purpose: We would like to clarify the relation between the learning curve and geometry of neuro-manifolds when the set of true parameters is an analytic set (the set defined as zeros of an analytic function). Methods: The following topics will be discussed. (1) Information Geometry of Singular Neuro-Manifold. (2) Theory of Order Statistic. (3) Weak Convergence and Empirical Prcess. (4) Conic Singularities. (5) Algebraic Geometry and Resolution of Singuralities. (6) Zeta function of Kullback Information and Prior. ******************************************* From Pete.Roper at nih.gov Wed Mar 14 11:24:48 2001 From: Pete.Roper at nih.gov (Pete Roper) Date: Wed, 14 Mar 2001 11:24:48 -0500 Subject: Postdoctoral Fellow Position Available Message-ID: <3AAF9B4E.438C@nih.gov> Postdoctoral Fellow Vacancy in the Laboratory of Cellular and Synaptic Neurophysiology; National Institute of Child Health and Human Development A position is available for a postdoctoral fellow within the Laboratory of Dr Chris J. McBain, Laboratory of Cellular and Synaptic Neurophysiology at the NICHD. Candidates should have experience in computational neurosciences with particular emphasis on the modeling of cortical neuronal networks. The candidate will work with an outstanding team of electrophysiologists to model excitatory and inhibitory synaptic circuitry and voltage gated potassium channels expressed on local circuit inhibitory neurons and principal neurons within the CA3 subfield of the hippocampal formation (see Atzori and McBain, Nature Neurosci. 3:791, 2000; Toth et al JNeurosci 15: 8279; Toth and McBain Nature Neurosci 1; 572, 1998). Candidates should have less than five years postdoctoral experience. Interested applicants should email Dr Chris McBain at chrismcb at codon.nih.gov -- ------------------------------ Dr P Roper Mailing Address: Postdoctoral Fellow BSA Building, Suite 350 Mathematical Research Branch, 9190 Rockville Pike National Institute of Diabetes and Bethesda, MD 20892-2690 Digestive and Kidney Diseases, USA National Institutes of Health Bethesda, MD 20892. E?mail: pete.roper at nih.gov Phone: (301) 496-9644 Fax: (301) 402-0535 From mike at stats.gla.ac.uk Wed Mar 14 09:38:09 2001 From: mike at stats.gla.ac.uk (Mike Titterington) Date: Wed, 14 Mar 2001 14:38:09 +0000 (GMT) Subject: JOB : Postdoctoral research assistant at Glasgow Message-ID: A POSTDOCTORAL RESEARCH OPPORTUNITY AT THE UNIVERSITY OF GLASGOW (Ref 129/01) APPROXIMATE APPROACHES TO LIKELIHOOD AND BAYESIAN STATISTICAL INFERENCE IN INCOMPLETE-DATA PROBLEMS Applications are sought for a Post-Doctoral Research Assistantship (IA) at the University of Glasgow. The appointee will be based in the Department of Statistics, under the direction of Professor D. M. Titterington. The appointment will be made on the standard IA scale of 16775 - 25213 and will be for up to 36 months, starting on October 1, 2001, or as soon as possible thereafter. The post is funded by the UK Engineering and Physical Sciences Research Council (EPSRC). The research concerns an area in which algorithms such as the EM algorithm and Markov chain Monte Carlo procedures can suffer from high computational complexity and the appointee will investigate and extend variational and other approximations that are currently available in the statistical and neural-computing literatures. Applications, supported by full curriculum vitae and the names of three referees, should be sent to Professor D. M. Titterington, Department of Statistics, University of Glasgow, Glasgow G12 8QQ, Scotland, to arrive not later than April 13, 2001. Informal enquiries and requests for further particulars can be made to mike at stats.gla.ac.uk. ================================= D.M. Titterington, Department of Statistics, University of Glasgow, Glasgow G12 8QQ, Scotland, UK. mike at stats.gla.ac.uk Tel (44)-141-330-5022 Fax (44)-141-330-4814 http://www.stats.gla.ac.uk/~mike From inmanh at cogs.susx.ac.uk Thu Mar 15 11:17:53 2001 From: inmanh at cogs.susx.ac.uk (Inman Harvey) Date: Thu, 15 Mar 2001 16:17:53 +0000 Subject: MSc in Evolutionary and Adaptive Systems Message-ID: <3AB0EB31.9653C7EA@cogs.susx.ac.uk> The Evolutionary and Adaptive Systems (EASy) group at the University of Sussex is probably the largest such multidisciplinary research group in the world, working on a wide range of topics where Computer Science and Complex Systems and AI and Artificial Life swap ideas with Biology. We have over 30 active researchers at doctoral and postdoctoral level, plus a similar number of Masters students. The EASy MSc is a one year course with 2 terms of coursework followed by a major supervised project in a relevant area. If you are interested you can find out about us on http://www.cogs.susx.ac.uk/lab/adapt/index.html **Expanded funding** Due to its success over the last 4 years the UK EPSRC has now dramatically increased the financial support for this course. The number of studentships available has increased very significantly, awarded competitively: UK students tuition + living expenses, other EU students tuition fees only. Other international and local students not awarded studentships welcome on self-funding basis, part-time option over 2 years also available. The expanded funding also includes new facilities and new courses to be added. Faculty directly involved in the course include Dr Inman Harvey - artificial evolution, evolutionary robotics, artificial life Dr Phil Husbands - evolutionary computation, GasNets for robotics Dr Ezequiel Di Paolo - evolving collective behaviour, homeostasis, autopoiesis Dr Adrian Thompson - evolvable hardware, evolutionary electronics Other faculty here at Sussex in associated areas include Prof John Maynard Smith (Evolutionary theory) Prof Maggie Boden (Philosophy, Creativity, Artificial Life) Prof Andy Clark (Philosophy) Prof Tom Collett (Ant and bee navigational behaviour) Prof Mick O'Shea (Neuroscience) For further information and applications contact Linda Thompson COGS University of Sussex Brighton BN1 9QH pgadmissions at cogs.susx.ac.uk http://www.cogs.susx.ac.uk/lab/adapt/index.html =============================== -- Inman Harvey >> Evolutionary and Adaptive Systems Group << >> COGS, Univ. of Sussex, Brighton BN1 9QH, UK << inmanh at cogs.susx.ac.uk >> http://www.cogs.susx.ac.uk/users/inmanh/ << From apbraga at cpdee.ufmg.br Fri Mar 16 12:52:08 2001 From: apbraga at cpdee.ufmg.br (Antonio de Padua Braga) Date: Fri, 16 Mar 2001 14:52:08 -0300 Subject: Multi-objective optimization/bias-variance Message-ID: <3AB252C8.AFC7FDC5@cpdee.ufmg.br> Dear Connectionists, The following paper has just been published by Neurocomputing. The idea of the paper is to balance the error of the training set and the norm of the weight vectors with a multi-objective optimization approach to avoid over-fitting. Copies are available on request. We apologize in advance for any multiple postings that may be received. *********************************************************************** Improving generalization of MLPs with multi-objective optimization Teixeira, R.A., Braga, A.P., Takahashi, R.H.C. And Rezende, R. Neurocomputing. Volume 35, pages 189-194. ABSTRACT This paper presents a new learning scheme for improving generalization of Multilayer Perceptrons (MLPs). The algorithm uses a multi-objective optimization approach to balance between the error of the training data and the norm of network weight vectors to avoid over-fitting. The results are compared with Support Vector Machines (SVMs) and standard backpropagation. *********************************************************************** -- Prof. Antonio de Padua Braga, Depto. Eng. Eletronica, Campus da UFMG (Pampulha), C. P. 209, 30.161-970, Belo Horizonte, MG, Brazil Tel:+55 31 4994869, Fax:+55 31 4994850, Email:apbraga at cpdee.ufmg.br, http://www.cpdee.ufmg.br/~apbraga From lunga at ifi.unizh.ch Sun Mar 18 12:34:22 2001 From: lunga at ifi.unizh.ch (Max Lungarella) Date: 18 Mar 2001 18:34:22 +0100 Subject: EDEC2001 (2nd call) Message-ID: <3AB4F19E.C10C2391@ifi.unizh.ch> Note: Our apologies if you receive this posting more than once. CALL FOR CONTRIBUTIONS EDEC2001 - EMERGENCE AND DEVELOPMENT OF EMBODIED COGNITION (2nd call) Symposium at the 3rd International Conference on Cognitive Science August 27-31, 2001, Beijing, China SCOPE The objective of the symposium is to bring together researchers from cognitive science, psychology, engineering, robotics, artificial intelligence, philosophy, and related fields so as to further our understanding of embodiment and development, in particular their mutual relationship. Ultimately, the goal is to understand the emergence of high-level cognition of an organism interacting with its physical and social environment over extended periods of time. FOCUS The symposium will focus on research that explicitly takes embodiment into account, either at the level of computational models, or real-world devices, and on empirical work that explicitly attempts to explain the relation of developmental processes to embodiment. Finally, contributions giving a broad and novel philosophical or methodological view on embodied cognition are welcome. CONTRIBUTIONS Contributions are solicited from the following areas (but not restricted to this list): - Cognitive developmental robotics - Neural mechanisms of learning and development (e.g. neural networks, statistical, information theoretic) - Development of sensory and motor systems - Perception-action coupling, sensory-motor coordination - Categorization, object exploration - Communication and Social interaction - Methodologies - Debates and philosophical issues (e.g. constructivism vs. selectionism, nature nurture, scalability, symbol grounding) ORGANIZATION This will be a one-day symposium with a number of talks, with a lot of room for discussion, and a poster session. The poster session will be over a cocktail to ensure relaxed atmosphere. FORM OF CONTRIBUTION Contributions can be in the form of full papers, or abstracts for posters. Full papers must be 5 pages maximum, with fonts at least 12 pt. For posters and work in progress, please submit a one-page abstract. Accepted contributions will be published in the proceedings of the ICCS 2001. GUIDELINES FOR ABSTRACT/PAPER SUBMISSION The contribution should be submitted electronically to lunga at ifi.unizh.ch (Max Lungarella) in pdf or MS Word files. IMPORTANT DATES Submission deadline (full-length paper and abstract): April 30, 2001 Notice of acceptance: May 31, 2001 PROGRAM COMMITTEE Rolf Pfeifer (AI Lab, University of Zurich, chair) Max Lungarella (AI Lab, University of Zurich, Switzerland) Yasuo Kuniyoshi (Electrotechnical Laboratory, Tsukuba, Japan) Olaf Sporns (Department of Psychology, Indiana University, Bloomington, IN, USA) Giorgio Metta (LIRA-Lab, University of Genova, Italy) Giulio Sandini (LIRA-Lab, University of Genova, Italy) Rafael Nunez (University of Fribourgh, Switzerland, and University of California, Berkeley) From geoff at giccs.georgetown.edu Sun Mar 18 22:09:01 2001 From: geoff at giccs.georgetown.edu (Geoff Goodhill) Date: Sun, 18 Mar 2001 22:09:01 -0500 Subject: Papers available: topographic mappings Message-ID: <200103190309.WAA30535@brecker.giccs.georgetown.edu> The following 3 recent papers about topographic mappings (both natural and artificial) can now be downloaded from www.giccs.georgetown.edu/~geoff/pubs.html Haese, K. & Goodhill, G.J. (2001). Auto-SOM: Recursive Parameter Estimation for Guidance of Self-Organizing Feature Maps. Neural Computation , 13, 595-619. Goodhill, G.J. & Cimponeriu, A. (2000). Analysis of the elastic net model applied to the formation of ocular dominance and orientation columns. Network, 11, 153-168. Goodhill, G.J. (2000). Dating behavior of the retinal ganglion cell. Neuron, 25, 501-503. Regards, Geoff Geoffrey J Goodhill, PhD Assistant Professor, Department of Neuroscience & Georgetown Institute for Cognitive and Computational Sciences Georgetown University Medical Center 3900 Reservoir Road NW, Washington DC 20007 Tel: (202) 687 6889, Fax: (202) 687 0617 Email: geoff at giccs.georgetown.edu Homepage: www.giccs.georgetown.edu/labs/cns From ahirose at info-dev.rcast.u-tokyo.ac.jp Mon Mar 19 02:13:32 2001 From: ahirose at info-dev.rcast.u-tokyo.ac.jp (ahirose@info-dev.rcast.u-tokyo.ac.jp) Date: Mon, 19 Mar 2001 16:13:32 +0900 (JST) Subject: CFP KES'2001 Special Session: Complex-valued neural networks Message-ID: <200103190713.QAA08147@info-dev.rcast.u-tokyo.ac.jp> Dear Connectionists, The Special Session "Complex-valued neural networks and the applications" is now being organized for the International Conference KES'2001 Osaka, September 6-8, 2001 (http://www.bton.ac.uk/kes/kes2001/) In these years the complex-valued networks expand the application fields in optoelectronic imaging, remote sensing, quantum neural devices and systems, spatiotemporal analysis of physiological neural systems as well as artificial neural processing. The potentially wide applicability yields new aspects of theories required for novel or more effective functions and mechanisms. The Special Session aims at an inspiring discussion on the recent progress and the future development. A few more speakers will be invited in addition to: Hiroyuki Aoki (Tokyo National College of Technology) George Georgiou (California State University) Akira Hirose (Tokyo University) Iku Nemoto (Tokyo Denki University) Tohru Nitta (Electrotechnical Laboratory) If you are interested, please make contact by March 31 with the Session Chair: Akira Hirose, University of Tokyo ahirose at ee.t.u-tokyo.ac.jp -- From tomsimpson at hotmail.com Mon Mar 19 13:31:08 2001 From: tomsimpson at hotmail.com (Tom Simpson) Date: Mon, 19 Mar 2001 18:31:08 -0000 Subject: New Issue of Connexions Message-ID: APOLOGIES FOR CROSS-POSTING This is to announce the release of a new issue of Connexions - an online journal of issues in Philosophy and Cognitive Science. In this issue we have two papers: 1. 'Connectionism is Nothing but Control Theory' by Asim Roy Abstract: The paper shows that connectionist systems are based on standard control theory notions. It does so by first examining the notion of a controller in any system and then establishing that many of the simpler connectionist learning methods - like back-propagation, adaptive resonance theory (ART), reduced coulomb energy (RCE), and radial basis function (RBF) - use controllers in them. The paper shows the existence of controllers in these methods and shows that these controllers may be (1) within the learning system itself, (2) outside of the learning system or (3) a combination of the two. By logical extension, more complex connectionist systems, ones that use these simpler learning mechanisms within them, are also in turn using controllers and are therefore based on control theoretic concepts. The analysis of these connectionist systems is performed purely on a logical basis, by a logical analysis of their conceptual structure, and has nothing to do with their implementation, whether by use of neurocomputers or other kinds of computers. In general, this analysis implies that control theoretic notions are applicable to developing systems similar to the brain and refutes the claim in connectionism that their methods do not embody standard control theory concepts, that they have introduced a qualitatively new set of concepts and mechanisms and 2. 'On the Origin of Symbols' by Thomas E. Dickins. Abstract: Through a synthesis of Dunbars (1993, 1996) and Gomezs (1998a, b) hypotheses on the evolution of language a further hypothesis about the origins of symbolic communication is made that relies upon simple learning. The aim of this speculation is to propose a specific origin story for symbolic communication and to marry behaviourist and cognitivist concerns about language. It is argued that this order of approach will enforce a more realistic parsimony on future models of language evolution. We also have a review, by Tom Stafford, of 'Darwin Among the Machines' by George Dyson. Connexions is run by the Department of Philosophy at the University of Sheffield, England, and can be found at http://www.shef.ac.uk/uni/academic/N-Q/phil/connex/ Connexions is an on-line journal for all issues in the Philosophy of Cognitive Science. We primarily publish work-in-progress with the intention of exposing this work to friendly but rigorous criticism, and, more broadly, of generating debate on issues within cognitive science. Comments and replies in response to work published in the journal are encouraged, and can be mailed to the editor or posted on our discussion list Connex-L. Contributions for subsequent publications are also welcome. Details on how to join the list, post messages and submit articles and reviews can be found on the site, together with previous issues and further information about Connexions. Thank you for your time. Tom Simpson Editor tomsimpson at hotmail.com ----------------------------------------------------------- Tom Simpson Department of Philosophy University of Sheffield Sheffield S10 2TN http://www.shef.ac.uk/misc/personal/pip98ts _________________________________________________________________________ Get Your Private, Free E-mail from MSN Hotmail at http://www.hotmail.com. From a.burkitt at medoto.unimelb.edu.au Tue Mar 20 21:59:08 2001 From: a.burkitt at medoto.unimelb.edu.au (Anthony BURKITT) Date: Wed, 21 Mar 2001 13:59:08 +1100 Subject: Three preprints on integrate-and-fire neuron models Message-ID: <2DE39EB1B64BD31184C100C00D006EF9A16A38@mail.medoto.unimelb.edu.au> Dear Connectionists, I would like to announce the availability of three papers that have been accepted for publication and are of potential interest to people working on integrate-and-fire neurons. Electronic copies are available at the site: http://www.medoto.unimelb.edu.au/people/burkitta/Pubs.html or contact me at a.burkitt at medoto.unimelb.edu.au Feedback and comments are naturally very welcome. Cheers, Tony Burkitt =============================================== Synchronization of the neural response to noisy periodic synaptic input A. N. Burkitt and G. M. Clark The timing information contained in the response of a neuron to noisy periodic synaptic input is analyzed for the leaky integrate-and-fire neural model. We address the question of the relationship between the timing of the synaptic inputs and the output spikes. This requires an analysis of the interspike interval distribution of the output spikes, which is obtained in the Gaussian approximation. The conditional output spike density in response to noisy periodic input is evaluated as a function of the initial phase of the inputs. This enables the phase transition matrix to be calculated, which relates the phase at which the output spike is generated to the initial phase of the inputs. The interspike interval histogram and the period histogram for the neural response to ongoing periodic input are then evaluated by using the leading eigenvector of this phase transition matrix. The synchronization index of the output spikes is found to increase sharply as the inputs become synchronized. This enhancement of synchronization is most pronounced for large numbers of inputs and lower frequencies of modulation, and also for rates of input near the critical input rate. However, the mutual information between the input phase of the stimulus and the timing of output spikes is found to decrease at low input rates as the number of inputs increases. The results show close agreement with those obtained from numerical simulations for large numbers of inputs. http://www.medoto.unimelb.edu.au/people/burkitta/periodic.ps.zip Accepted for publication in Neural Computation (to appear). =============================================== Shot noise in the leaky integrate-and-fire neuron N. Hohn and A. N. Burkitt We study the influence of noise on the transmission of temporal information by a leaky integrate-and-fire neuron using the theory of shot noise. The model includes a finite number of synapses and has a membrane potential variance de facto modulated by the input signal. The phenomenon of stochastic resonance in spiking neurons is analytically exhibited using an inhomogeneous Poisson process model of the spike trains, and links with the traditional Ornstein-Uhlenbeck process obtained by a diffusion approximation are given. It is shown that the modulated membrane potential variance inherent to the model gives better signal processing capabilities than the diffusion approximation. http://www.medoto.unimelb.edu.au/people/burkitta/article_PRE_ES7178.ps.zip Published in Phys. Rev. E 63, 031902. =============================================== Balanced neurons: Analysis of leaky integrate-and-fire neurons with reversal potentials A. N. Burkitt A new technique is presented for analyzing leaky integrate-and-fire neurons that incorporates reversal potentials, which impose a biologically realistic lower bound to the membrane potential. The time distribution of the synaptic inputs is modeled as a Poisson process. The analysis is carried out in the Gaussian approximation, which comparison with numerical simulations confirms is most accurate in the limit of a large number of inputs. The hypothesis that the observed variability in the spike times of cortical neurons is caused by a balance of excitatory and inhibitory synaptic inputs is supported by the results for the coefficient of variation of the interspike intervals. It's value decreases with both increasing numbers and amplitude of inputs, and is consistently lower than 1.0 over a wide range of realistic parameter values. The dependence of the output spike rate upon the rate, number and amplitude of the synaptic inputs, as well as upon the value of the inhibitory reversal potential, are given. http://www.medoto.unimelb.edu.au/people/burkitta/balanced.ps.zip Accepted for publication in Biological Cybernetics (to appear). ====================ooOOOoo==================== Anthony N. Burkitt The Bionic Ear Institute 384-388 Albert Street East Melbourne, VIC 3002 Australia Email: a.burkitt at medoto.unimelb.edu.au http://www.medoto.unimelb.edu.au/people/burkitta Phone: +61 - 3 - 9283 7510 Fax: +61 - 3 - 9283 7518 =====================ooOOOoo=================== From bengio at idiap.ch Wed Mar 21 05:26:42 2001 From: bengio at idiap.ch (Samy Bengio) Date: Wed, 21 Mar 2001 11:26:42 +0100 (MET) Subject: Several open positions in speech, vision, machine learning at IDIAP Message-ID: SEVERAL OPEN POSITIONS IN SPEECH, COMPUTER VISION, MACHINE LEARNING, AND MULTIMODAL INTERFACES The Dalle Molle Institute for Perceptual Artificial Intelligence (IDIAP, http://www.idiap.ch) is a semi-private research institute affiliated with the Swiss Federal Institute of Technology at Lausanne (EPFL) and the University of Geneva. Located in Martigny (Valais, CH), IDIAP is partly funded by the Swiss Federal Government, the State of Valais, and the City of Martigny, and is involved in numerous national and international (European) projects. IDIAP is mainly carrying research and development in the fields of speech and speaker recognition, computer vision, and machine learning, and is recognized as a university level research laboratory, involving permanent senior scientists, postdocs, and PhD students (usually awarded an EPFL degree). IDIAP currently numbers around 35-40 scientists. IDIAP, in close collaboration with EPFL (mainly with the Signal Processing Laboratory of Prof. Murat Kunt, http://ltswww.epfl.ch) will soon be the "Leading House" of a large Research Network (National Center of Competence in Research) on "Interactive Multimodal Information Management", which also opens up several new opportunities for long term positions at different levels. In view of the resulting (present and future) growth of the Institute, IDIAP currently welcomes applications of talented candidates at all levels with expertise or strong interest in the fields of speech processing, computer vision, machine learning, and multimodal interaction. The open positions include: management and senior positions (including one scientific deputy director and one speech processing group leader), project leaders, postdocs, and PhD students. Two EPFL tenure track positions at the Assistant Professor (with most of the research responsibilities located at IDIAP) are also available. Preference will be given to candidates with experience in one or several of the following areas: signal processing, statistical pattern recognition (typically applied to speech and scene analysis), neural networks, hidden Markov models, speech and speaker recognition, computer vision, human/computer interaction (dialog). Senior and postdoc candidates should also have a proven record of high quality research and publications. All applicants should be experienced in C/C++ programming and familiar with the Unix environment; they should also be able to speak and write in English (and be willing to learn French). LOCATION: IDIAP is located in the town of Martigny (http://www.martigny.ch) in Valais, a scenic region in the South of Switzerland, surrounded by the highest mountains of Europe, and offering exciting recreational activities (including hiking, climbing and skiing), as well as varied cultural activities. It is also within close proximity to Montreux, Lausanne (EPFL) and Lake Geneva, and centrally located for travel to other parts of Europe. PROSPECTIVE CANDIDATES should send their detailed CV, together with a motivation letter and 3 reference letters, to: Prof. Herv? Bourlard, Director of IDIAP, P.O. Box 592, Simplon, 4 CH-1920 Martigny Switzerland Email: bourlard at idiap.ch Phone: +41-27-721.77.20; Fax: +41-27-721.77.12 ----- Samy Bengio Research Director. Machine Learning Group Leader. IDIAP, CP 592, rue du Simplon 4, 1920 Martigny, Switzerland. tel: +41 27 721 77 39, fax: +41 27 721 77 12. mailto:bengio at idiap.ch, http://www.idiap.ch/~bengio From J.Hancock at cs.rhul.ac.uk Wed Mar 21 07:58:13 2001 From: J.Hancock at cs.rhul.ac.uk (John Hancock) Date: Wed, 21 Mar 2001 12:58:13 +0000 Subject: Bioinformatics Postdoc, Univerity of London UK Message-ID: Postdoctoral Research Assistant Bioinformatics Research Group, Department of Computer Science, Royal Holloway University of London We are looking for a highly motivated individual to join the bioinformatics research group in the Department. The aim of the project, funded by the BBSRC's bioinformatics initiative, is to apply machine learning techniques (such as the Support Vector Machine) to the identification of coding regions, gene promoters and other sequence features in genomic sequences, with special emphasis on plant genomes. This is a three year appointment commencing as soon as possible. The Department of Computer Science at Royal Holloway has a leading position in the study of theory and practice of machine learning and in particular the development of the Support Vector learning technique and other kernel-based techniques. The Department is located on Royal Holloway's pleasant campus (see http://www.rhul.ac.uk/) in Englefield Green, Surrey, close to London. The successful candidate will work with John Hancock, Alex Gammerman and Victor Solovyev in collaboration with Peter Bramley in the School of Biological Sciences. We are looking for someone with a strong background in computational biology, particularly as applied to genomic sequence analysis, and preferably also some programming skills and knowledge of machine learning. The post would also be ideal for someone with a computer science or maths background who is looking to move into bioinformatics. Salary is in the range 18,909 to 22,599 UK pounds per annum inclusive of London Allowance. Information on the Department may be found at www.cs.rhul.ac.uk/. Informal enquiries may be directed to John Hancock (J.Hancock at cs.rhul.ac.uk). See http://www.cs.rhul.ac.uk/home/jhancock/ for more information on the project and John Hancock's research. Further details and an application form can be obtained from The Personnel Office, Royal Holloway, University of London, Egham, Surrey TW20 0EX; fax: +44 (0)1784 473527; tel: +44 (0)1784 414241; email Sue.Clarke at rhul.ac.uk Please quote reference KB/1717. The closing date for applications is 23rd April 2001. We positively welcome applications from all sections of the community. *************************************************** Dr John M. Hancock Reader in Computational Biology, Director of Graduate Studies, Department of Computer Science, Royal Holloway University of London, Egham, Surrey TW20 0EX, U.K. *************************************************** Phone: +44 (01784) 414 256 Fax: +44 (01784) 439 786 Mobile: +44 (077) 9046 6302 E-mail: J.Hancock at cs.rhul.ac.uk WWW: http://www.cs.rhul.ac.uk/home/jhancock/ *************************************************** From ishii at is.aist-nara.ac.jp Thu Mar 22 01:22:02 2001 From: ishii at is.aist-nara.ac.jp (Shin Ishii) Date: Thu, 22 Mar 2001 15:22:02 +0900 Subject: Preprint available Message-ID: <200103220622.PAA16405@dec127.aist-nara.ac.jp> Dear connectionists, We are pleased to inform you that the following preprint is now available on our Web page: http://www.aist-nara.ac.jp/~ishii/Papers/nc_psn.ps.gz Comments and suggestion would be greatly appreciated. Authors: Ken-ichi Amemori and Shin Ishii Title: Gaussian process approach to spiking neurons for inhomogeneous Poisson inputs (Neural Computation, to appear) Keywords: spiking neuron, stochastic process, Markov-Gaussian process, inhomogeneous Poisson input, first passage time density Abstract: This article presents a new theoretical framework to consider the dynamics of a stochastic spiking neuron model with general membrane response to input spike. We assume that the input spikes obey an inhomogeneous Poisson process. The stochastic process of the membrane potential then becomes a Gaussian process. When a general type of the membrane response is assumed, the stochastic process becomes a Markov-Gaussian process. We present a calculation method for the membrane potential density and the firing probability density. Our new formulation is the extension of the existing formulation based on diffusion approximation. Although the single Markov assumption of the diffusion approximation simplifies the stochastic process analysis, the calculation is inaccurate when the stochastic process involves a multiple Markov property. We find that the variation of the shape of the membrane response, which has often been ignored in existing stochastic process studies, significantly affects the firing probability. Our approach can consider the reset effect, which has been difficult to be dealt with by analysis based on the first passage time density. --- Shin Ishii Nara Institute of Science and Technology http://www.aist-nara.ac.jp/~ishii/ From erik at bbf.uia.ac.be Thu Mar 22 10:09:44 2001 From: erik at bbf.uia.ac.be (Erik De Schutter) Date: Thu, 22 Mar 2001 16:09:44 +0100 Subject: European short-term fellowships in neuroinformatics and computational neuroscience Message-ID: The EU Thematic Network Computational Neuroscience and Neuroinformatics offers short-term fellowships to nationals of the EU and associated countries for visits to laboratories in the EU or associated countries. Fellowships cover travel and accommodation costs for a visit of 2 to 12 weeks. Applications are evaluated once every month on a competitive basis. Fellowships can be awarded within three months of the application. Award per fellowship varies between 600 to 2000 =A4. More information and electronic application is available at http:www.neuroinf.org Prof. E. De Schutter University of Antwerp, Belgium coordinator of the Thematic Network Computational Neuroscience and Neuroinformatics http://www.bbf.uia.ac.be From john at cs.rhul.ac.uk Fri Mar 23 12:29:22 2001 From: john at cs.rhul.ac.uk (John Shawe-Taylor) Date: Fri, 23 Mar 2001 17:29:22 +0000 (GMT) Subject: Applications of learning to text and images Message-ID: Expressions of interest are invited for the following workshop: NeuroCOLT/KerMIT workshop on Applications of Learning to Text and Images Windsor, 30 April - 2 May 2001 Cumberland Lodge, Windsor, UK http://www.neurocolt.org/cumberland2001.html REGISTRATION DEADLINE: APRIL 3rd The continuous growth of information repositories in digital format (eg the World Wide Web) makes it necessary to develop entirely new - and automatic - ways of accessing it. Learning methods are one of the most promising research directions. This workshop will bring together experts from information retrieval, text categorization, machine vision, and learning theory. The aim is to provide an understanding of the state of the art, and of the open problems to which learning technology could be applied. The list of invited speakers is on the web site - they will be giving two hour tutorial style talks. For registration and constantly updated information, pls click here: http://www.neurocolt.org/cumberland2001.html The cost will be approximately 100 UKP per day inclusive of bed and full board. ************************************************************** John Shawe-Taylor J.Shawe-Taylor at cs.rhul.ac.uk Dept of Computer Science, Royal Holloway, University of London Phone: +44 1784 443430 Fax: +44 1784 439786 ************************************************************** From shultz at psych.mcgill.ca Fri Mar 23 10:55:53 2001 From: shultz at psych.mcgill.ca (Thomas R. Shultz) Date: Fri, 23 Mar 2001 10:55:53 -0500 Subject: recent papers on cognitive and language development Message-ID: <4.3.1.0.20010323104943.00a7cc30@127.0.0.1> Recent papers on cognitive and language development that may be of interest to readers of this list Shultz, T. R., & Bale, A. C. (2001, in press). Neural network simulation of infant familiarization to artificial sentences: Rule-like behavior without explicit rules and variables. Infancy. A fundamental issue in cognitive science is whether human cognitive processing is better explained by symbolic rules or by sub-symbolic neural networks. A recent study of infant familiarization to sentences in an artificial language claims to have produced data that can only be explained by symbolic rule learning and not by unstructured neural networks. Here we present successful unstructured neural network simulations of the infant data, showing that these data do not uniquely support a rule-based account. In contrast to other simulations of these data, the present simulations cover more aspects of the data with fewer assumptions about prior knowledge and training, using a more realistic coding scheme based on sonority of phonemes. The networks show exponential decreases in attention to a repeated sentence pattern, more recovery to novel sentences inconsistent with the familiar pattern than to novel sentences consistent with the familiar pattern, occasional familiarity preferences, more recovery to consistent novel sentences than to familiarized sentences, and extrapolative generalization outside the range of the training patterns. A variety of predictions suggest the utility of the model in guiding future psychological work. The evidence, from these and other simulations, supports the view that unstructured neural networks can account for the existing infant data. ================ Sirois, S., Buckingham, D., & Shultz, T. R. (2000). Artificial grammar learning by infants: An auto-associator perspective. Developmental Science, 4, 442-456. This paper reviews a recent article suggesting infants use a system of algebraic rules to learn an artificial grammar (Marcus, Vijayan, Bandi Rao, & Vishton, 1999). In three reported experiments, infants exhibited increased responding to auditory strings that violated the pattern of elements they were habituated to. We argue that a perceptual interpretation is more parsimonious, as well as more consistent with a broad array of habituation data. We report successful neural network simulations that implement a lower-level interpretation and capture the empirical regularities reported by Marcus and colleagues (1999). The discussion puts the simulation results in the context of the broader debate about interpreting infant habituation. Other neural network models of habituation in general, and of the Marcus et al. (1999) task specifically, are discussed. ================ Buckingham, D., & Shultz, T. R. (2000). The developmental course of distance, time, and velocity concepts: A generative connectionist model. Journal of Cognition and Development, 1, 305-345. Connectionist simulations of children's acquisition of distance (d), time (t), and velocity (v) concepts using a generative algorithm, cascade-correlation, are reported. Rules that correlated most highly with network responses during training were consistent with the developmental course of children's concepts. Networks integrated the defining dimensions of the concepts first by identity rules (e.g., v = d), then additive rules (e.g., v = d t), and finally multiplicative rules (e.g., v = d / t). The results are discussed in terms of similarity to children's development, the contribution of connectionism to the study of cognitive development, contrasts with alternate models, and directions for future research. ================ Oshima-Takane, Y., Takane, Y., & Shultz, T. R. (1999). The learning of first and second pronouns in English: Network models and analysis. Journal of Child Language, 26, 545-575. Although most English-speaking children master the correct use of first and second person pronouns by three years, some children show persistent reversal errors in which they refer to themselves as you and to others as me. Recently, such differences have been attributed to the relative availability of overheard speech during the learning process. The present study tested this proposal with feed-forward neural networks learning these pronouns. Network learning speed and analysis of their knowledge representations confirmed the importance of exposure to shifting reference provided by overheard speech. Errorless pronoun learning was linked to the amount of overheard speech, interactions with a greater number of speakers, and prior knowledge of the basic-level kind PERSON. Preprints and reprints can be found at http://www.psych.mcgill.ca/perpg/fac/shultz/default.htm Cheers, Tom -------------------------------------------------------- Thomas R. Shultz, Professor, Department of Psychology, McGill University, 1205 Penfield Ave., Montreal, Quebec, Canada H3A 1B1. E-mail: shultz at psych.mcgill.ca Updated 23 March 2001: http://www.psych.mcgill.ca/perpg/fac/shultz/default.htm Phone: 514 398-6139 Fax: 514 398-4896 -------------------------------------------------------- From vdavidsanchez at earthlink.net Sun Mar 25 19:22:55 2001 From: vdavidsanchez at earthlink.net (V. David Sanchez A.) Date: Sun, 25 Mar 2001 16:22:55 -0800 Subject: CFP NEUROCOMPUTING Special Issue on Neural Pattern Recognition Message-ID: <3ABE8BDF.3EAE761F@earthlink.net> CALL FOR PAPERS NEUROCOMPUTING An International Journal published by Elsevier Science B.V., vol. 36-41, 24 issues, in 2001 ISNN 0925-2312, URL: http://www.elsevier.nl/locate/neucom Special Issue on Neural Pattern Recognition Paper Submission Deadline: June 30th, 2001 Neural Pattern Recognition has been traditionally considered as an alternative approach to pattern recognition, different from the conventional statistical and structural approaches. There have been much success in practical pattern recognition applications using neural networks including Multilayer Perceptrons, Radial Basis Functions, and Self-Organizing Maps among others. Lately, there has been strong convergence between neural and statistical classifiers, exemplified by algorithms like Bayesian learning, the Expectation Maximization (EM) algorithm for mixture modeling, Support Vector Machines (SVM), or Independent Component Analysis (ICA) for feature extraction to mention just a few. The Neurocomputing journal invites original contributions for the forthcoming special issue on Neural Pattern Recognition from a broad scope of areas. Some topics relevant to this special issue include, but are not restricted to: -- Theoretical foundations -- Neural architectures for pattern recognition -- Hybrid recognition systems -- Relationship between neural and statistical classifiers -- Unsupervised learning in feature extraction -- Use of spiking neurons in pattern recognition -- Key applications including classification, clustering, data mining, bioinformatics, speech recognition, character recognition, 3-D object recognition, etc. Please send two hardcopies of the manuscript before June 30th, 2001, to: V. David Sanchez A., Neurocomputing - Editor in Chief - Advanced Computational Intelligent Systems P.O. Box 60130, Pasadena, CA 91116-6130 U.S.A. Street address: 1149 Wotkyns Drive Pasadena, CA 91103 Fax: +1-626-793-5120 Email: vdavidsanchez at earthlink.net including abstract, keywords, a cover page containing the title and author names, corresponding author name's complete address including telephone, fax, and email address, and clear indication to be a submission to the Special Issue on Neural Pattern Recognition. Guest Editors Krzysztof J. Cios University of Colorado at Denver Computer Science and Engineering Department U.S.A. Phone: (303) 556-4314 Fax: (303)556-8369 Email: kcios at carbon.cudenver.edu Kunihiko Fukushima The University of Electro-Communications Dept. of Information and Communication Engineering 1-5-1, Chofugaoka, Chyofu-shi, Tokyo 182-8585 Japan Fax +81-424-43-5291 Email: fukushima at ice.uec.ac.jp Erkki Oja, Helsinki University of Technology. Laboratory of Computer and Information Science P.O. Box 5400, 02015 HUT Finland Phone +358-9-4513265 Fax. +358-9-4513277 Email Erkki.Oja at hut.fi V. David Sanchez A., Neurocomputing - Editor in Chief - Advanced Computational Intelligent Systems P.O. Box 60130 Pasadena, CA 91116-6130 U.S.A. Fax: +1-626-793-5120 Email: vdavidsanchez at earthlink.net From steve at cns.bu.edu Tue Mar 27 11:26:18 2001 From: steve at cns.bu.edu (Stephen Grossberg) Date: Tue, 27 Mar 2001 08:26:18 -0800 Subject: frequency-dependent synaptic depression, Hebbian pairing, and synchrony in cortical pyramidal neurons Message-ID: The following article is now available at http://www.cns.bu.edu/Profiles/Grossberg in HTML, PDF, and Gzipped Postscript. Okatan, M. and Grossberg, S. Frequency-Dependent Synaptic Potentiation, Depression, and Spike Timing Induced by Hebbian Pairing in Cortical Pyramidal Neurons. Neural Networks ABSTRACT: Experiments by Markram and Tsodyks (1996) have suggested that Hebbian pairing in cortical pyramidal neurons potentiates or depresses the transmission of a subsequent pre-synaptic spike train at steady-state depending on whether the spike train is of low frequency or high frequency, respectively. The frequency above which pairing induced a significant decrease in steady-state synaptic efficacy was as low as about 20 Hz and this value depends on such synaptic properties as probability of release and time constant of recovery from short-term synaptic depression. These characteristics of cortical synapses have not yet been fully explained by neural models, notably the decreased steady-state synaptic efficacy at high pre-synaptic firing rates. This article suggests that this decrease in synaptic efficacy in cortical synapses was not observed at steady-state, but rather during a transition period preceding it whose duration is frequency-dependent. It is shown that the time taken to reach steady-state may be frequency-dependent, and may take considerably longer to occur at high than low frequencies. As a result, the pairing-induced decrease in synaptic efficacy at high pre-synaptic firing rates helps to localize the firing of the post-synaptic neuron to a short time interval following the onset of high frequency pre-synaptic spike trains. This effect may "speed up the time scale" in response to high frequency bursts of spikes, and may contribute to rapid synchronization of spike firing across cortical cells that are bound together by associatively learned connections. Key Words: synaptic potentiation, synaptic depression, frequency-dependent synaptic plasticity, cortical pyramidal cells, Hebbian pairing, cortical synchronization From jose at psychology.rutgers.edu Tue Mar 27 11:14:04 2001 From: jose at psychology.rutgers.edu (Stephen J. Hanson) Date: Tue, 27 Mar 2001 11:14:04 -0500 Subject: UNIX/LINUX GURUs! with Neural Net or Cognitive Neuroscience Interests.. Message-ID: <3AC0BC4C.1030803@kreizler.rutgers.edu> Departmental Systems Adminstration/UNIX/LINUX IMMEDIATE OPENING-- 04/01/01 PSYCHOLOGY DEPARTMENT-RUTGERS UNIVERSITY--Newark Campus--We are searching for an individual who can adminster the computing resources of the Psychology Department at Rutgers University (Newark). Resources include a network of SUN (Enterprise Server) /LINUX workstations, PCs and Macs, printers, voice mail system and various devices (scanners, projecters etc..). The individual will be responsible for installing and debugging software, and various routine system administration activites. Some proportion of time will be spent in research involving Cognitive Science especially related to Connectionist networks (or Neural Networks and Cognitive Neuroscience. Familiarity with C programming, UNIX system internals (BSD, System V, Solaris, Linux) and Windows (XX, NT) and local area networks running TCP/IP is required. Image processing or graphics programing experience are pluses. Candidates should possess either a BS/MS in Computer Science, Cognitive Science, AI or other relevant fields or equivalent experience. We are located 15-20 minutes from Downtown Manhattan and are within minutes of the NJPAC in the heart of Newark. Salary is competitive and will be dependent upon qualifications and experience. Rutgers University is an equal opportunity affirmative action employer. Please send snail mail resumes and references to Stephen J. Hanson Department of Psychology 101 Warren Street Rutgers University Newark, New Jersey, 07102 (Preferred) Direct email inquiries or resumes to: jose at psychology.rutgers.edu Please indicate on SUBJECT Line: SYS ADM as a keyword. From watrous at scr.siemens.com Tue Mar 27 13:07:11 2001 From: watrous at scr.siemens.com (Raymond L Watrous) Date: Tue, 27 Mar 2001 13:07:11 -0500 (EST) Subject: Career opportunity in signal processing R&D Message-ID: <200103271807.NAA03975@mt0.scr.siemens.com> CAREER OPPORTUNITY Senior Engineer Zargis Medical Corp. Princeton, NJ Zargis Medical Corp. has an immediate opening for a senior signal processing engineer. About Zargis Medical Zargis Medical Corp. is a newly formed company that develops and markets advanced diagnostic decision support products and services for primary care physicians and other healthcare professionals. Zargis Medical was formed through a joint investment by Siemens Corporate Research, a subsidiary of Siemens Corporation, and SPEEDUS.COM, Inc. (Nasdaq: SPDE). Job Description The senior engineer will provide leadership in developing advanced signal processing algorithms for the analysis and interpretation of heart sounds. The senior engineer will be involved in clinical data acquisition, product design, software development, quality assurance, and clinical trials. Qualifications The candidate should have an advanced degree in biomedical or electrical engineering, and a minimum of 5 years of experience in research and development. The candidate should possess superior mathematical ability and strong statistical signal processing skills. Familiarity with time-frequency methods, adaptive filtering, feature extraction, pattern matching, nonlinear optimization and hidden Markov models are essential. A background in biomedical signal processing or speech recognition would be an asset. The candidate should have demonstrated effectiveness in algorithm development, algorithm analysis and validation, technical communications and project management. Compensation Zargis Medical offers a competitive compensation and benefits plan. Zargis provides a congenial work environment and an outstanding career opportunity in an emerging technology company with excellent growth potential. Next Step Please forward a copy of your resume to: Raymond Watrous, Ph.D. Chief Technology Officer Zargis Medical Corp. 755 College Road East Princeton, NJ 08540 TEL: (609) 734-6596 FAX: (609) 734-6565 Email: careers at zargis.com From ismip00 at ee.usyd.edu.au Sat Mar 31 20:17:07 2001 From: ismip00 at ee.usyd.edu.au (ismip) Date: Sun, 1 Apr 2001 11:17:07 +1000 Subject: TNN Special Issue Final Call for Papers Message-ID: <004901c0ba49$7b794d20$471b4e81@pc.ee.usyd.edu.au> We apologize if you receive the message multiple times. Final Call for Papers IEEE Transaction on Neural Networks Special Issue on Intelligent Multimedia Processing Human communication is intrinsically multimodal. With the advances of technology, modern communication systems will also become more and more multimodal. Hence, multimedia technologies represent new ground for research interactions among a variety of media such as speech, audio, image, video, text and graphics. Future multimedia technologies will need to handle information with an increasing level of intelligence, i.e., automatic recognition and interpretation of multimodal signals. This is particularly emphasized in MPEG-7 which focuses on the 'multimedia content description interface'. The description shall be associated with the content itself to facilitate fast and effective searching for all the media. Specifically, the MPEG-7 research domain will cover techniques for content-based indexing and retrieval: pattern recognition, face detection/recognition, and fusion of multimodality. Intelligent multimedia processing shares three fundamental goals with biological systems: a) Universal data processing engine for multimodal signals; b) Multimodality; and c) Unsupervised clustering and/or supervised learning by examples. Because of these features, neural networks are attractive candidates for intelligent multimedia processing and recent activity in the area is a proof of this fact. The main attribute of neural computing is its adaptive learning capability, which enables interpretations of possible variations of a same object or pattern, e.g., with respect to scale, orientation, and perspective. Moreover, they are able to accurately approximate unknown systems based on sparse sets of noisy data. Certain neural models also effectively incorporate statistical signal processing and optimization techniques. In addition, spatial/temporal neural structures and hierarchical models are promising for multirate, multiresolution multimedia processing. As a result, many successful applications of neural networks in intelligent multimedia processing, sometimes combined with fuzzy systems and evolutionary computing, have been reported. The possible topics for the special issue include, but are not limited to, the following: * Neural networks (including BSS and ICA) and other computational intelligence models, learning paradigms, and architectures for multimedia processing. * Intelligent multimedia processing architectures. * Multimedia/multichannel data fusion. * Multimodal representation and information retrieval: Applications in hyperlinking of multimedia objects, query and search of multimedia information including intelligent web agents, 3D object representation and motion tracking, image sequence generation and animation. * Human-computer interaction and communications: face recognition, lip-reading analysis, facial expression and emotion categorization, interactive human-machine vision, speech recognition, speaker recognition, gesture analysis and recognition, auditory/visual scene analysis, and multimodal interaction. * Multimedia data analysis and visualization: texture, color, content, etc. * Intelligent network control of audio/video streams in multimedia networking applications. Original, previously-unpublished research articles as well as state-of-the-art tutorial papers will be considered. Authors should follow the IEEE TNN manuscript format described in the Information for Authors, which can be found on the inside back cover of any issue of TNN. Prospective authors are invited to submit papers to the website: http://eivind.imm.dtu.dk/tnn. The following schedule will apply: Manuscript submission: April 15, 2001 Acceptance notification: July 31, 2001 Final manuscripts due: October 30, 2001 Publication: January 2002 Guest Editors: Tulay Adali, Ling Guan, Dept of CSEE School of Electrical & Information Eng. Univ of Maryland, Baltimore County The University of Sydney Baltimore, MD 21250 Sydney, NSW 2000 Australia Jan Larsen Shigeru Katagiri Dept of Mathematical Modelling ATR Technical University of Denmark 2-2 Hikaridai 2800 Lyngby Seika-cho, Soraku-gun Denmark Kyoto 619-02 Japan Jose Principe Dept of Electrical & Computer Eng University of Florida Gainesville, FL 32611