From eero at cns.nyu.edu Fri Dec 1 01:02:27 2006 From: eero at cns.nyu.edu (Eero Simoncelli) Date: Fri, 1 Dec 2006 01:02:27 -0500 (EST) Subject: Connectionists: COSYNE 2007: Abstract Submission and Registration Open Message-ID: <200612010602.kB162R008320@calaf.cns.nyu.edu> *************************************************************************** Computational and Sytems Neuroscience (COSYNE) MAIN MEETING WORKSHOPS Feb 22-25, 2007 Feb 26-27, 2007 Salt Lake City, UTAH The Canyons, UTAH http://cosyne.org/ *************************************************************************** ABSTRACT SUBMISSION and REGISTRATION are now open at http://cosyne.org/ IMPORTANT DATES --------------- * Abstract submission deadline: 15-Dec-06 * Complete schedule release: 25-Jan-07 * Early on-line registration ends: 31-Jan-07 The annual COSYNE meeting provides an inclusive forum for the discussion of experimental and theoretical approaches to problems in systems neuroscience. The meeting is expected to draw about 350-400 researchers from a wide variety of disciplines. Topics include but are not limited to: neural coding; natural scene statistics; dendritic computation; neural basis of persistent activity; nonlinear receptive field mapping; representations of time and sequence; reward systems; synaptic plasticity; map formation and plasticity; population coding; attention; computation with spiking networks. The MAIN MEETING, held in Salt Lake City, will be single-track, and will consist of both oral and poster sessions. Some oral presentations will be invited, while others will be drawn from short submitted abstracts. Poster presentations will be drawn from submitted abstracts. Invited speakers for this year include: * Ehud Ahissar (Weizmann Institute) * Richard Andersen (Caltech) * Ed Callaway (Salk Institute) * Paul Glimcher (NYU) * Michael Goldberg (Columbia) * Judith Hirsch (USC) * Mitsuo Kawato (ATR) * Eric Knudsen (Stanford) * Mike Lewicki (CMU) * Zhaoping Li (UCL) * Dan Margoliash (U Chicago) * Bruce McNaughton (U Arizona) * Bartlett Mel (USC) * Sheila Nirenberg (Cornell) * Mike Shadlen (U Washington) The WORKSHOPS will be at the Canyons ski resort nearby, and will offer parallel sessions for more in-depth discussion of specialized topics. Preliminary workshop topics are as follows: 1. How silent/active is the brain? 2. Hippocampal and entorhinal coding across species (2 days) 3. Emerging information-theoretic measures and methods in neuroscience 4. Neurally plausible statistical inference 5. Functional requirements of a visual theory 6. Conserved functions of the basal ganglia circuit 7. What role does spike synchrony or correlation play in sensory processing? 8. Asking why - normative models in neuroscience 9. Quantitative analysis of shape representation in mid and higher level visual areas 10. Random matrix theory and neural networks 11. Motor control 12. Decision making For further information, please consult http://cosyne.org or send email to cosyne at rochester.edu From ilafiete at caltech.edu Fri Dec 1 02:28:18 2006 From: ilafiete at caltech.edu (Ila Fiete) Date: Thu, 30 Nov 2006 23:28:18 -0800 Subject: Connectionists: computational neuroscience postdoc position available at Caltech Message-ID: <456FD992.3070107@caltech.edu> Apologies for multiple postings if you are subscribed to several groups. Announcement of postdoctoral position in computational/theoretical neuroscience at the Center for Neural Systems, Caltech -- Available: 1 fully-funded postdoctoral position to collaborate with me and the greater systems neuroscience community at Caltech. I'm looking for individuals who are interested in understanding neural codes and dynamics in neural networks that underlie complex behaviors (for example, but not limited to, singing in songbirds, navigation using hippocampal and entorhinal circuits in rodents, and odor coding in insects). Work will be carried out in collaboration with relevant systems neuroscience experimental labs at Caltech and elsewhere. Candidates should have a strong quantitative background in Physics, Mathematics, Engineering, or Computer Science. The start date is flexible, but with a preference for begining in Spring 2007. The initial appointment will be for 1 year with a possibility of an extension. Ability to program in Matlab and C, and some knowledge of neuroscience a plus, but not necessary. If interested, please contact me (ilafiete at caltech.edu) with a copy of your cv or resume, a statement of research interestes, and 3 letters of reference. For more information, please look at http://www.klab.caltech.edu/~ila/ and http://www.cns.caltech.edu/ Best regards, Ila Fiete From terry at salk.edu Sun Dec 3 03:44:51 2006 From: terry at salk.edu (Terry Sejnowski) Date: Sun, 03 Dec 2006 00:44:51 -0800 Subject: Connectionists: NEURAL COMPUTATION December 2006 In-Reply-To: Message-ID: Neural Computation - Contents - Volume 18, Number 12 - December 1, 2006 Article Short-term Synaptic Plasticity Can Enhance Weak Signal Detectability in Non-renewal Spike Trains Niklas Lüdtke and Mark E. Nelson Notes On the use of Analytic Expressions for the Voltage Distribution to Analyze Intracellular Recordings M. Rudolph and A. Destexhe A Distributed Computing Tool for Generating Neural Simulation Databases Robert J. Calin-Jageman and Paul S. Katz Kernel Least Squares Models Using Updates of the Pseudoinverse E. Andelic, M. Schafföner, M. Katz, S.E. Krüger, A. Wendemuth Learning Tetris Using the Noisy Cross-Entropy Method István Szita and András Lörincz Letters A Spiking Neuron Model of Cortical Correlates of Sensorineural Hearing Loss: Spontaneous Firing, Synchrony and Tinnitus Melissa Dominguez, Suzanna Becker, Ian Bruce and Heather Read Event-driven Simulation Scheme for Spiking Neural Networks Using Look-Up Tables to Characterize Neuronal Dynamics Eduardo Ros, Richard Carrillo, Eva M. Ortigosa, Boris Barbour, and Rodrigo Agís On the Computational Power of Threshold Circuits with Sparse Activity Kei Uchizawa, Wolfgang Maass, and Rodney Douglas Parameter Space Structure of Continuous-Time Recurrent Neural Networks Randall D. Beer Bifurcation Analysis of Jansen's Neural Mass Model François Grimbert and Olivier Faugeras A Model for Perceptual Averaging and Stochastic Bi-stable Behaviour, and the Role of Voluntary Control Ansgar R. Koene A Unifying View of Wiener and Volterra Theory and Polynomial Kernel Regression Matthias O. Franz and Bernhard Schölkopf An Upper Bound on the Minimum Number of Monomials Required to Separate Dichotomies of {-1, 1}^n Erhan Oztop ----- ON-LINE - http://neco.mitpress.org/ SUBSCRIPTIONS - 2006 - VOLUME 18 - 12 ISSUES Electronic only USA Canada* Others USA Canada* Student/Retired $60 $64.20 $114 $54 $57.78 Individual $100 $107.00 $154 $90 $96.30 Institution $730 $781.10 $784 $657 $702.99 * 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 B.Kappen at science.ru.nl Fri Dec 1 03:17:09 2006 From: B.Kappen at science.ru.nl (Bert Kappen) Date: Fri, 01 Dec 2006 08:17:09 -0000 Subject: Connectionists: postdoc at SNN and Keygene Message-ID: Postdoc position available at Keygene and SNN, the Netherlands. Keygene is a biotechnology company that carries out innovative research to advance commercial plant breeding by using a number of proprietary genetic technologies and know-how. To expedite creation of new varieties by its customers, Keygenes main focus is the analysis, creation and exploitation of genetic variation. SNN is a research group at Nijmegen University dedicated to fundamental research in the areas of machine learning and computational neuroscience. Specific topics are Bayesian networks, approximate inference methods, time-series modeling, bio-informatics, expert systems, stochastic control and collaborative decision making. The group consists currently of 8 researchers and three programmers. Keygene and SNN are collaborating within the area of in silico breeding. Combining favorable alleles by crossing existing well-characterized plant lines with a minimum of efforts is a major challenge in marker-assisted crop breeding. An optimal crossing scenario has to be found that guarantees ending up with the desired combination of alleles in a single new plant line with a predefined probability. At Keygene, some initial research steps have been taken to develop algorithms that avoid calculating all possibilities but still end up with close to optimal solutions. The current project will built further on these initial methods, further improving these and enabling putting constraints on the chosen scenario, defined by demands from the breeding practice. For this novel research topic, we are looking for an ambitious and excellent postdoc quantitative genetics. The requirement for the postdoc position is a PhD and a good track record in the field of bioinformatics or machine learning. The postdoc position is full-time for a period of 3 years and will spend his time equally at both Keygene and SNN. The postdoc will be employed by Keygene as part of the Keygene postdoc program (see attached). For more information see www.snn.kun.nl/nijmegen or contact Prof. dr. H.J. Kappen (b.kappen at science.ru.nl, tlf. +31 24 3614241). -- Prof. dr. H.J. Kappen SNN Radboud University Nijmegen URL: www.snn.ru.nl/~bertk The Netherlands tel: +31 24 3614241 fax: +31 24 3541435 B.Kappen at science.ru.nl mobile: +31 6 520 78 210 From rhaschke at techfak.uni-bielefeld.de Fri Dec 1 05:32:34 2006 From: rhaschke at techfak.uni-bielefeld.de (Robert Haschke) Date: Fri, 01 Dec 2006 10:32:34 -0000 Subject: Connectionists: WSOM07 - Announcement and Call for Papers Message-ID: <457004BA.10205@techfak.uni-bielefeld.de> *********************************************************** C A L L F O R P A P E R S 6th Int. Workshop on Self-Organizing Maps (WSOM 2007) September 3rd - 6th 2007, Bielefeld, Germany http://www.wsom07.org *********************************************************** -- Important Dates -- Manuscript Submission 15th April 2007 Acceptance Notification 8th June 2007 Camera Ready Submission 1st July 2007 Early Registration 15th July 2007 Conference 3rd - 6th Sep 2007 -- About The Workshop -- The 6th International Workshop on Self-Organizing Maps will bring together researchers and practicioners in the field of self-organizing systems, with a particular emphasis on neural networks and self-organizing maps. It will highlight key advances in these and related fields. It is the sixth conference in a series of bi-annual international conferences started with WSOM'97 in Helsinki. The programme committee cordially invites you to attend WSOM 2007 and submit papers on all aspects of self-organizing systems including: * Self-organization * Unsupervised learning (including PCA/NLPCA, ICA/NLICA, Principal Curves/Surfaces) * Signal processing, image processing and vision * Robotics and Intelligent Systems * Data visualization, mining and sonification * Bioinformatics * Text and document analysis * Financial analysis * Time-series analysis * Theory and extensions * Optimization * Hardware and architecture Papers will be peer reviewed by an international program committee according to the criteria pertinence, scientific quality, impact, generality, and innovation. High-quality submissions will be selected for oral or poster presentation during the conference. The conference programme will include invited talks as well as oral and poster presentations of selected papers. It will be enriched by official workshops to allow practitioners and scientists to discuss focussed topics in an atmosphere that fosters the exchange of ideas. Besides the scientific goals of the conference, the organisers strive to provide an accessible meeting by keeping expenses for participants affordable, particularly for young researchers. -- Programme Committee -- Prof. Estevez, Pablo (University of Chile) Dr. Flanagan, Adrian (Nokia Research Center, Finland) Prof. Furukawa, Tetsuo (Kyushu Institute of Technology, Japan) Prof. Fyfe, Colin (University of Paisley, UK) Prof. Hammer, Barbara (University of Osnabr?ck, Germany) Prof. Van Hulle, Marc (KU Leuven, Belgium) Prof. Kaski, Samuel (Helsinki University of Tech., Finland) Prof. Lampinen, Jouko (Helsinki University of Tech., Finland) Prof. Martinetz, Thomas (University of L?beck, Germany) Prof. Miikkulainen, Risto (University of Texas at Austin, USA) Prof. Pag?s, Gilles (University of Paris, France) Dr. Polani, Daniel (University of Hertfordshire, UK) Prof. Simula, Olli (Helsinki University of Tech., Finland) Prof. Torras, Carme (University of Catalonia, Spain) Prof. Ultsch, Alfred (University of Marburg, Germany) Prof. Verleysen, Michel (Universit? catholique de Louvain, Belgium) Prof. Villmann, Thomas (University Leipzig, Germany) Prof. Wermter, Stefan (University of Sunderland, UK) -- Special Issue -- The best papers of the conference will be selected for a reviewed special issue of a renowned scientific journal. More specific information will be provided soon. Please watch our website. -- Workshops -- The WSOM07 conference committee cordially invites members of the community to submit proposals for workshops to be held in conjunction with the main conference. Workshops should facilitate a lively discussion about emerging topics or specific aspects in the area of self-organizing systems. If you have any questions regarding workshop topics or organisation, please feel free to contact the conference office via email at info at wsom07.org or have a look at the conference website for further information. -- Instructions for Authors -- Prior to the submission deadline, please connect to the conference web server to enter the title, author names and addresses, abstract and keyword. The paper will then be assigned a paper identification number. Full papers must be submitted according to the style guidelines published on the conference website. For further and updated information, please visit the conference website http://www.wsom07.org -- Organizing Committee -- Honorary Conference Chair Prof. Teuvo Kohonen (Helsinki University of Technology) General Conference Chair Helge Ritter (Bielefeld University, Germany) Steering Committee Nigel Allinson (UMIST, Manchester, UK) Marie Cottrell (SAMOS-MATISSE, Paris 1, France) Erkki Oja (Helsinki Univ. of Tech., Finland) Helge Ritter (Bielefeld University, Germany) Takeshi Yamakawa (Kyushu Inst. of Tech., Japan) Hujun Yin (UMIST, Manchester, UK) Local Chair Robert Haschke (Bielefeld University, Germany) -- Enquiry and Information -- WSOM 2007 Conference Secretariat Neuroinformatics Group Bielefeld University, Germany http://www.wsom07.org/contact.html Email: info at wsom07.org -- Int. Workshop on Self-Organizing Maps 3rd-6th September 2007 http://www.wsom07.org Bielefeld, Germany Robert Haschke, Bielefeld University, Germany From M.Casey at surrey.ac.uk Mon Dec 4 03:45:43 2006 From: M.Casey at surrey.ac.uk (M.Casey@surrey.ac.uk) Date: Mon, 04 Dec 2006 08:45:43 -0000 Subject: Connectionists: Information Fusion Journal Special Issue Message-ID: Call for papers for a special issue of the journal Information Fusion on "Biologically Inspired Information Fusion" Deadline for Submission: January 30, 2007 Full details at http://www.elsevier.com/authored_subject_sections/P05/CFP/IFJ_CFP_IDS.pd f Bringing together research on the biology and psychology of multi-sensory processing, computational neuroscience and theoretical work on mechanisms for combining different information sources. Topics appropriate for this special issue include, but are not limited to: * Biologically inspired fusion schemes * Adaptive information fusion which emphasize biological motivations * Biologically inspired fusion in robotics * Multimodal integration: * Modeling combined sensory processing * Including, but not exclusively, combining vision, audition, olfaction, taste or touch * Combining artificial and biological sensors * Attention or emotional biasing on sensory processing * Biologically motivated applications of multi-sensor integration Our understanding of both natural and artificial cognitive systems is an exciting area of research that is developing into a multi-disciplinary subject with the potential for significant impact on science, engineering and society in general. There is considerable interest in how our understanding of natural systems may help us to apply biological strategies to artificial systems. Of particular interest is our understanding of how to build adaptive information fusion systems by combining knowledge from different domains. In natural systems, the integration of sensory information is learnt at an early stage of development. Therefore, through a better understanding of the structures and processes involved in this natural adaptive integration, we may be able to construct a truly artificial multi-sensory processing system. Here then, psychological and physiological knowledge of multisensory processing, and particularly the low level influence that different modalities have on one another, can be used to build upon existing theoretical work on computational mechanisms, such as self organization and the combination of multiple neural networks, to build systems that can fuse together different information sources. These themes were recently discussed at an International Workshop on Biologically Inspired Information Fusion. As well as presenting the state-of-the-art on multi-sensory processing and information fusion from the life and physical sciences, the workshop provided a forum for researchers to discuss priorities for developing this multi-disciplinary area. This special issue of Information Fusion is therefore aimed at following up from these discussions by focusing on the highlighted priorities, whilst also providing an opportunity for the wider dissemination of relevant themes. For this special issue, papers should either have a biological motivation and/or inspiration, or otherwise be of biological relevance and interest. Manuscripts should make the biological dimension explicit. Information Fusion related papers lacking this dimension are to be submitted to a regular issue of the journal. Manuscripts (which should be original and not previously published or presented even in a more or less similar form under any other forum) covering biologically inspired information fusion methods and their applications as well as the theories and algorithms developed to address these applications are invited. Contributions should be described in sufficient detail to be reproducible on the basis of the material presented in the paper. Manuscripts should be submitted electronically online at http://ees.elsevier.com/inffus. The corresponding author will have to create a user profile if one has not been established before at Elsevier. Simultaneously, please also send without fail an electronic copy (PDF format preferred), to the Guest Editors listed below. Please identify clearly that the submission is meant for this special issue. Guest Editors Dr Matthew Casey, Department of Computing, University of Surrey, UK, m.casey at surrey.ac.uk Professor Robert Damper, School of Electronics and Computer Science, University of Southampton, UK, rid at ecs.soton.ac.uk Further information can be found at: http://www.cs.surrey.ac.uk/people/academic/M.Casey/biif2006.html http://www.elsevier.com/locate/inffus From fsommer at berkeley.edu Tue Dec 5 03:17:22 2006 From: fsommer at berkeley.edu (Fritz Sommer) Date: Tue, 05 Dec 2006 08:17:22 -0000 Subject: Connectionists: GRADUATE PROGRAM IN NEUROSCIENCE - UC BERKELEY Message-ID: GRADUATE PROGRAM IN NEUROSCIENCE - UC BERKELEY ** Application deadline: December 15, 2006 ** The Graduate Program in Neuroscience at UC Berkeley is currently accepting applications for admission for the 2007-2008 academic year. There are numerous opportunities for students interested in focusing on computational and theoretical approaches within the context of an interdisciplinary neuroscience training program. Faculty supporting this area include: Martin Banks - Visual space perception, psychophysics, virtual reality Jose Carmena - Brain-machine interfaces, sensorimotor control, learning Yang Dan - Information processing in thalamus and cortex Mike DeWeese - Auditory processing, computational models Jack Gallant - Neural mechanisms of visual form perception and attention Tom Griffiths - Computational models of cognition Stanley Klein - Computational models of spatial vision, psychophysics Harold Lecar - Theoretical biophysics, network models Bruno Olshausen - Models of visual cortex, scene analysis Fritz Sommer - Models of associative memory, sensory processing Frederic Theunissen - Neural mechanisms of complex sound recognition Frank Werblin - Information processing in the retina Faculty in other programs also pursuing computational/theoretical approaches to neuroscience questions include Michael Gastpar, EECS - Neural coding, information theory Jitendra Malik, EECS - Models of early vision and object recognition Alva Noe, Philosophy - Theories of perception and sensorimotor loops Lokendra Shastri, ICSI - Models of episodic memory in hippocampus Bin Yu, Statistics - Neural coding, image statistics In addition, the newly established Redwood Center for Theoretical Neuroscience provides a central workspace for theoreticians and organizes a weekly seminar series, workshops, and hosts visiting scholars. See http://redwood.berkeley.edu For further information and details of the application processs see http://neuroscience.berkeley.edu/grad.php and please note the application deadline above. From brad_love at mail.utexas.edu Wed Dec 6 01:02:21 2006 From: brad_love at mail.utexas.edu (Brad Love) Date: Wed, 06 Dec 2006 06:02:21 -0000 Subject: Connectionists: Prizes for Best Conference Papers Message-ID: Five prizes of $1000 (USD) each are available to top papers submitted to the 2007 Annual Conference of the Cognitive Science Society, which will be held in Nashville from August 1st to 4th. The submission deadline is February 1st. The conference website is http:// csep.psyc.memphis.edu/cogsci07/ Four prizes worth $1,000 (USD) each will be awarded for the best paper submissions that involve computational cognitive modeling. The four separate prizes will represent the best modeling work in the respective areas of: perception/action, language, higher-level cognition, and applied cognition. Any form of computational cognitive modeling relevant to cognitive science will be eligible, including (but not limited to) connectionism, symbolic, Bayesian, dynamic systems, or various hybrids. Additionally, the Marr Prize, named in honor of the late David Marr, will be given to the best student paper at the conference. More details about these prizes can be found at http:// csep.psyc.memphis.edu/cogsci07/prizes.htm#comprizesec ********************************************************************* Bradley C. Love The University of Texas at Austin brad_love at mail.utexas.edu 1 University Station A8000 http://love.psy.utexas.edu/ Austin, TX 78712-0187 phone: (512) 232-5732 fax: (512) 471-5935 ********************************************************************* From online at gavrila.net Wed Dec 6 10:33:34 2006 From: online at gavrila.net (online@gavrila.net) Date: Wed, 06 Dec 2006 15:33:34 -0000 Subject: Connectionists: JOB: PhD Position "Advanced Pedestrian Recognition" - DaimlerChrysler R&D, Germany Message-ID: <1165419206.4576e2c64f927@www.domainfactory-webmail.de> PhD Position: Advanced Pedestrian Recognition The Machine Perception Department of DaimlerChrysler R&D in Ulm, south-Germany, has currently an opening for a Ph.D. student in the area of computer vision. The open position concerns our ongoing research on the challenging (but very interesting) topic of vision-based pedestrian detection from a moving vehicle. Pedestrians are arguably the most vulnerable traffic participants; more than 6000 pedestrians are killed and 150.000 injured yearly in the EU. Especially children are at risk. A number of separate vision modules have been developed in our group to handle this problem: stereo- and radar-based obstacle detection, optical-flow based detection of independently moving objects, shape-based object detection, intensity-based pattern classification and tracking. So far, these modules have only been partially combined, and mostly in loose and rather ad-hoc fashion. The primary focus of the Ph.D. research lies therefore in how to better integrate these modules in a high-performance multi-cue object detection system, using sound probabilistic principles. A secondary focus of the Ph.D. research is the pattern classification module. Here we are interested in automatic and interactive techniques to sample new training data in order to improve classifier performance (active learning). Experiments are to be performed both off-line, on unique large and prorietary data sets, and on-line, on-board our Mercedes-Benz vehicle demonstrator, on the test track and in real urban traffic. Prospective applicants should demonstrate an above-average academic record and a solid background in applied mathematics and computer science. Experience in computer vision and pattern recognition is a definite plus. The successful applicant will furthermore be motivated and able to conduct research independently, as well as collaborate effectively with other members of a research team. A certain affinity towards real-time vision systems and towards turning complex techniques into real-world practice is essential. Language skills: good in English, basic in German. Funding for the Ph.D. student position is available immediately, for a period of 3-4 years. External academic supervision to be provided by Univ. of Mannheim (Prof. Dr. C. Schn?rr). Prospective applicants please email cover letter (explaining suitability), resume, copy of academic record, references, and any supporting materials (e.g. MS Thesis, publications, GRE scores) to Prof. Dr. Dariu Gavrila Machine Perception, GRE/AP DaimlerChrysler Research & Technology Wilhelm Runge St. 11, 89081 Ulm, Germany [my-first-name].[my-last-name]@DaimlerChrysler.com For more background information, see http://www.gavrila.net/Computer_Vision/Research/Pedestrian_Detection/pedestrian_detection.html From Mayank_Mehta at brown.edu Wed Dec 6 15:20:52 2006 From: Mayank_Mehta at brown.edu (Mehta, Mayank) Date: Wed, 06 Dec 2006 20:20:52 -0000 Subject: Connectionists: Graduate studies in Brain Sciences at Brown University References: <5.1.1.6.0.20051005153429.031869f0@pop.isc.cnrs.fr> Message-ID: Graduate studies in Brown University's Brain Science Program (BSP) provide interdisciplinary training that knits together cognitive, neural, and computational sciences. It emphasizes collaborative theoretical and experimental studies of the brain, from the molecular to the behavioral and cognitive levels. Interested students apply directly to one of the program's core departments and include a statement indicating their interest in the Brain Science Program. BSP provides educational and research opportunities that augment the department-based programs and prepares students for interdisciplinary research careers. For more information, contact brainscience at brown.edu or visit http://brainscience.brown.edu/apply.html --------------------- Mayank Mehta http://neurophysics.brown.edu From welazmeh at site.uottawa.ca Wed Dec 6 23:27:42 2006 From: welazmeh at site.uottawa.ca (William Elazmeh) Date: Thu, 07 Dec 2006 04:27:42 -0000 Subject: Connectionists: CFP - Workshop on Evaluation Methods for Machine Learning II Message-ID: <4266.127.0.0.1.1165465660.squirrel@127.0.0.1> Call for participation The AAAI-2007 Workshop on Evaluation Methods for Machine Learning II Description: ============ The purpose of this workshop is to continue the lively and interesting debate started last year at the AAAI 2006 workshop on Evaluation Methods for Machine learning. The previous workshop was successful on the following points: - It established that the current means of evaluating learning algorithms has some serious drawbacks. - It established that there are many important properties of algorithms that should be measured, requiring more than a single evaluation metric. - It established that algorithms must be tested under many different conditions. - It established that the UCI data sets do not reflect the variety of domains to which algorithms are applied in practice. For this year's workshop, we are intending to address, in a more specific fashion, some of the topics that were raised at last year's workshop and some new ones. Last year's participants are invited to submit papers reflecting the evolution of their views within the year, or new ideas. Researchers, or practitioners, interested in this issue who did not get a chance to submit a paper last year are invited to do so this year. Topics: ======= We invite position papers and technical papers addressing three main topics, and their subtopics: Evaluation Metrics: * The efficacy of existing evaluation metrics * The need for new metrics * Useful metrics from the other fields * Single number summaries versus curves * New performance visualization methods Statistical Issues: * Cross-validation vs. bootstrap * Bias vs. variance * Parametric vs. non-parametric * The power of tests * Sampling methods * Multiple comparisons Data Sets: * The (over) use of community repositories such as UCI * Concept Drift * Synthetic, or semi-synthetic, data sets Format and Attendance: ====================== The workshop will consist of: Invited talks: We are planning to have several invited speakers. Some, from outside of our research community, will be able to criticize our accepted practices from an external point of view. Some, from inside our community, will discuss how we could improve on our current practices. Panel Discussion: Our invited speakers will be asked to engage each other on the various issues surrounding the problem of evaluation in Machine Learning at the end of the workshop. The audience will be strongly encouraged to participate in the discussion. Presentations: The papers accepted to the workshop will be presented throughout the day between the various invited talks. Papers will be grouped by theme, in order to facilitate discussion at the end of each session. Workshop attendance is open to the public and is estimated at 20-25 attendees. Priority will be given to those active participants in the workshop (paper authors or speakers). Submission: =========== Authors are invited to submit papers on the topics outlined above or other related issues. Both position and technical reports will be considered for this workshop. To promote a lively event, with plenty of discussion, the organizers are very interested in papers taking strong positions on the issues listed above. Workshop papers should not exceed 6 pages using the AAAI Style. Submissions should be made electronically in PDF or Postscript format and should be sent (no later than April 1, 2007) by email to William Elazmeh: welazmeh at site.uottawa.ca Resubmissions and position papers (of at least 1 page) are welcome. Important Dates: ================ * Submission deadline: April 1, 2007 * Notification date: April 25, 2007 * Camera ready deadline: May 15, 2007. * The workshop date: July 22 or 23, 2007. * AAAI-07 conference is to be held on July 22-26, 2007, in Vancouver, British Columbia, Canada. Organizers: =========== Chris Drummond NRC Institute for Information Technology, Canada Email: chris.drummond at nrc-cnrc.gc.ca William Elazmeh (main contact) School of Information Technology and Engineering, University of Ottawa, 800 King Edward Ave., room 5-105, Ottawa, Ontario, K1N 6N5 Canada Telephone: +1 (613) 562 5800 Extension 6699 Fax.: +1 (613) 562 5664 Email: welazmeh at site.uottawa.ca Nathalie Japkowicz School of Information Technology and Engineering University of Ottawa, Canada Email: nat at site.uottawa.ca Sofus A. Macskassy Fetch Technologies United States of America Email: sofmac at fetch.com Program Committee: ================== Nitesh Chawla, University of Notre Dame, USA Peter Flach, University of Bristol, UK Alek Kotcz, Microsoft Research, USA Nicolas Lachiche, Louis Pasteur University, France Sylvain Letourneau, National Research Council, Canada Dragos Margineantu, The Boeing Company, Phantom Works, USA Stan Matwin, University of Ottawa, Canada Kiri Wagstaff, NASA Jet Propulsion Laboratory, USA Bianca Zadrozny, Fluminense Federal University, Brazil Workshop Website: ================= http://www.site.uottawa.ca/~welazmeh/conferences/AAAI-07/workshop/ From jf218 at cam.ac.uk Thu Dec 7 21:35:36 2006 From: jf218 at cam.ac.uk (Prof. J. Feng) Date: Fri, 08 Dec 2006 02:35:36 -0000 Subject: Connectionists: Reminder: International Spring School on Computational Neurobiology in Shanghai In-Reply-To: References: <5.1.1.6.0.20051005153429.031869f0@pop.isc.cnrs.fr> Message-ID: Dear All, The deadline is less than two weeks away. the best Jianfeng Computational Neurobiology Spring School in Shanghai March 19th to 28th, 2007 http://www.dcs.warwick.ac.uk/~feng/school.html CALL FOR PARTICIPATION The first International Spring School on Computational Neurobiology in Shanghai will be organized at the CAS-MPG Partner Institute for Computational Biology, Shanghai, PR China (http://www.icb.ac.cn). The school's speakers are listed below. Please, pass on this information to your interested colleagues and students. The course lasts for ten days. Lectures will be given by distinguished experts on topics across the full breadth of experimental and theoretical neuroscience. The first few days of the course introduces students to essential neurobiological concepts and to the most important techniques in modeling single cells, networks, and neural systems. Each student will be assigned to a tutor. In the afternoon, students are given lectures, practical training in modeling, and time to accomplish their report. At the end of the course, students are expected to submit a report of five pages. The course is designed for advanced graduate students and postdoctoral fellows in related disciplines. Students are expected to have a solid background in either (neuro)biology, mathematics, or computer science and, in addition, some basic knowledge in the other two fields as well as a keen interest in the subject matter. Students of any nationality can apply. A maximum of 30 students will be accepted. The course is free, every student has to cover costs for lodging and meals which will be rather moderate. Applications including a CV and an outline of the applicant's motivation for participation must be submitted electronically through the web site and should be accompanied by two letters of recommendation (also sent electronically). Applications will be assessed by the scientific committee, with selection being based on the following criteria: the scientific quality of the candidate (CV), the letters, and evidence that the course affords substantial benefit to the candidate's training. DEADLINE: Application BEFORE 20th Dec, 2006. Speakers at the Spring School 1. Larry Abbott (Columbia, USA) 2. Alexander Borst (Munich, Germany) 3. Yang Dan (Berkeley, USA) 4. Jianfeng Feng (Warwick, UK) 5. Juergen Jost (Leipzig, Germany) 6. Konrad P Koerding (MIT, USA) 7. Wolfgang Maass (Graz, Austria) 8. Hiroyuki Nakahara (RIKEN, Japan) 9. Miguel AL Nicolelis (Duke, USA) 10. Mu-Ming Poo (Berkeley, USA) 11. Ning Qian (Columbia, USA) 12. John Rinzel (New York, USA) 13. Helge Ritter (Bielefeld, Germany) 14. Xiao-jing Wang (Yale, USA) Scientific Committee: Andreas Dress (Shanghai), Jianfeng Feng (Warwick), Juergen Jost (Leipzig), Mu-Ming Poo (Berkeley), Helge Ritter (Bielefeld) From morgado_dias at clix.pt Sat Dec 9 11:57:10 2006 From: morgado_dias at clix.pt (morgado_dias@clix.pt) Date: Sat, 09 Dec 2006 16:57:10 -0000 Subject: Connectionists: Ph.D Scholarships at the University of Madeira Message-ID: <20061209165659.12867.qmail@maboque.srv.clix.pt> An embedded and charset-unspecified text was scrubbed... Name: not available Url: https://mailman.srv.cs.cmu.edu/mailman/private/connectionists/attachments/20061209/b7670afa/attachment.ksh From mark.plumbley at elec.qmul.ac.uk Thu Dec 14 12:06:33 2006 From: mark.plumbley at elec.qmul.ac.uk (Mark Plumbley) Date: Thu, 14 Dec 2006 17:06:33 -0000 Subject: Connectionists: ICA 2007: Intl. Conf. on Independent Component Analysis and Signal Separation: CALL FOR PAPERS Message-ID: <9D47A2D30B0BFB4C920786C25EC69345BDE94C@staff-mail.vpn.elec.qmul.ac.uk> Dear Connectionists, (apologies for any multiple receipts) ICA 2007 will be held in London, UK, next September. The Call for Papers is below: please forward to any colleagues who may interested in submitting or attending. This year the name of the conference has change slightly to reflect the fact that its scope is not only independent component analysis (ICA) and blind source separation (BSS), but also includes non-blind, semi-blind and model-based signal separation. If you have not been to ICA before, now might be the year to start! Best wishes, Mark Plumbley (General Chair, ICA 2007) ---------------------------------------------------------------------- *** FIRST CALL FOR PAPERS *** ICA 2007 7th International Conference on Independent Component Analysis and Signal Separation London, UK 9-12 September 2007 www.ica2007.org The 7th International Conference on Independent Component Analysis and Signal Separation (ICA 2007) will be held at Queen Mary, University of London, from Sunday 9 September to Wednesday 12 September 2007. Following the previous ICA conferences in Aussois (France), Helsinki (Finland), San Diego (CA, USA), Nara (Japan), Granada (Spain) and Charleston (SC, USA), this year the conference is organized and sponsored by the ICA Research Network, an EPSRC-funded network of researchers in over 25 UK institutions. The Conference will include invited talks as well as oral and poster presentations of refereed papers and special sessions. It will be organized in a single track and will be selective. Papers are solicited in all areas of independent component analysis and signal separation, including blind source separation (BSS), as well as semi-blind, non-blind, and model-based signal separation. Topics of interest include (but not limited to) the following: * Algorithms and Architectures: non-linear ICA, probabilistic models, sparse coding, linear & nonlinear models, convolutive & noisy models; * Theory: optimization, complex methods, time-frequency representations; * Applications: audio, bio-informatics, biomedical engineering, communications, finance, image processing, psychology; * Emerging Technologies: analogue and digital VLSI implementations, photonics; * Functional Neuroimaging: EEG, ERP, MEG, fMRI, applications in neuroscience; * Speech and Musical Audio: source separation, denoising, dereverberation, temporal models, computational auditory scene analysis (CASA), beamforming; * Visual and Sensory Processing: image processing and coding, image separation. All contributions must be original, and must not have been previously published, nor be under review for presentation elsewhere. Detailed instructions for submission to ICA?2007, including formatting instructions and templates, will be available from the conference website at http://www.ica2007.org Important Deadlines Special session proposals: 1 March 2007 Paper submission deadline: 23 March 2007 ---------------------------------------------------------------------- Dr Mark D Plumbley General Chair, ICA 2007 Department of Electronic Engineering Queen Mary University of London Mile End Road, London E1 4NS, UK Tel: +44 (0)20 7882 7518 Fax: +44 (0)20 7882 7997 Email: mark.plumbley at elec.qmul.ac.uk http://www.elec.qmul.ac.uk/people/markp/ From mail at mkaiser.de Fri Dec 15 12:20:03 2006 From: mail at mkaiser.de (Marcus Kaiser) Date: Fri, 15 Dec 2006 17:20:03 -0000 Subject: Connectionists: 3-year PostDoc position: Network analysis and Neuroinformatics Message-ID: <200612151720.kBFHK1F3026629@post.webmailer.de> Please forward the following position to potential candidates. Thanks! This appointment is part of the CARMEN initiative in Neuroinformatics which brings together a consortium of 19 investigators from 11 UK Universities with expertise in experimental Neuroscience, computing science and statistical data analysis. The initiative also involves neuroscientists in Germany, Japan and the USA, and a number of commercial companies who are interested in both the computing challenges and the way the project can help their studies of the brain. The research project involves the development of algorithms for the analysis and visualization of activity patterns in neural systems. These patterns could range from the firing rates of neurons from multi-electrode recordings to functional networks as yielded by functional imaging, or optical or electrophysiological recordings in EEG. In addition to implementing algorithms used in network analysis, new algorithms for detecting characteristic activity patterns will be developed. A key part of the role will be working with experimental neuroscientists to understand the nature of the data that will be analyzed. You will therefore need good communication skills, including the ability to give presentations on their work, and to interact with others from different disciplines. You will need to be able to write design documents and academic papers. You should ideally have, or be about to submit, a PhD in a theoretical subject such as computer science, physics, or mathematics. You will have knowledge and experience in one or more of: graph theory, network analysis, information theory, statistics, distributed computing or, the application of computing to a scientific domain. The project is jointly supervised by Dr Marcus Kaiser from Newcastle University and Dr Stephen Eglen from the University of Cambridge. Dr Kaiser is based at the Institute of Neuroscience in Newcastle which is one of the largest neuroscience groups in the UK with more than 100 independent investigators (http://www.ncl.ac.uk/ion/ ). Dr Eglen is based in the Cambridge Computational Biology Institute (CCBI), which is physically hosted in the Department of Applied Mathematics and Theoretical Physics at the University of Cambridge. One role of the CCBI is to increase interactions between experimentalists and theorists in the area of computational biology, of which computational neuroscience is one focus. Other ongoing research in my group includes the development, robustness, dynamics, and spatial organisation of cortical and neuronal networks (see Trends in Cognitive Sciences, 2004 and PLoS Computational Biology, 2006 articles available from my homepage). In particular we are interested in how activity spreads in cortical networks and how the brain recovers after lesions. Closing date for applications is: 15 January 2007 Informal enquiries about this position may be made to Dr Marcus Kaiser, m.kaiser at ncl.ac.uk Full details are available at http://www.ncl.ac.uk/vacancies (Job Ref: D1633R) More information about our groups can be found at http://www.biological-networks.org (Dr Marcus Kaiser) http://www.damtp.cam.ac.uk/user/sje30/ (Dr Stephen Eglen) -- Marcus Kaiser, Ph.D. RCUK Academic Fellow School of Computing Science University of Newcastle Claremont Tower Newcastle upon Tyne NE1 7RU, U.K. Phone: +44 191 222 8161 Fax: +44 191 222 8232 http://www.biological-networks.org/ From jvita_01 at uni-muenster.de Mon Dec 18 11:18:39 2006 From: jvita_01 at uni-muenster.de (Julien Vitay) Date: Mon, 18 Dec 2006 16:18:39 -0000 Subject: Connectionists: Workshop "From Computational Cognitive Neuroscience to Computer Vision" at ICVS2007 Message-ID: <6d49ed8d0612180818r15184a80wbb11dc8904e854e5@mail.gmail.com> As part of the 5th International Conference on Computer Vision Systems ( ICVS2007 ) that will take place from March 21st to 24th 2007 in Bielefeld (Germany), we ask for contributions for the workshop: *From Computational Cognitive Neuroscience to Computer Vision - CCNCV 2007* * * *Web-Page* http://www.icvs2007.org/ccncv.php *Description of the workshop* This workshop aims to foster the transfer of knowledge from the area of Computational Cognitive Neuroscience to Computer Vision. The seemingly simplicity of visual perception has periodically attracted machine vision scientists to build biologically inspired algorithms. Although such biologically motivated algorithms have not revolutionized Computer Vision, we believe that several reasons exist that motivate an ongoing exchange. In the recent years there has been much progress in the understanding of brain processes such as visual attention, object recognition, learning of complex feature detectors, decision making and high level cognitive control, to name a few. This work is yet often not suitable for real world applications, but it might lead to new developments, and in the long run, to better algorithms in machine vision. *General topics* We invite contributions from the area of Computational Cognitive Neuroscience that relate to Computer Vision. The contribution should emphasize those properties of neural models that potentially lead to applications in Computer Vision. We do not expect that these applications already exist. *Submissions (format, styles, etc.)* Abstracts of max 1 page (preferably PDF) should be submitted to fhamker at uni-muenster.de. The work should be novel but it can be already published. The primary criterion of acceptance is the relevance for the workshop. *Important Dates* *Submission of abstracts:* 15th January 2007 *Notification of acceptance:* 2nd February 2007 *Camera ready paper: * tbd. *Workshop on CCNCV:* Is planned to take place on Wednesday, March, 21st (a detailed schedule will be available here soon). * * *Workshop organizers, contact and e-mail for paper submission* *Fred Hamker * Psychological Institute II Otto Creutzfeldt Center for Cognitive and Behavioral Neuroscience Westf. Wilhelms-Universit?t M?nster E-mail: fhamker at uni-muenster.de Web: http://wwwpsy.uni-muenster.de/Psychologie.inst2/AELappe/personen/hamker.html *Julien Vitay * Psychological Institute II Westf. Wilhelms-Universit?t M?nster Web: http://julien.vitay.free.fr/ From esalinas at wfubmc.edu Mon Dec 18 14:43:11 2006 From: esalinas at wfubmc.edu (Emilio Salinas) Date: Mon, 18 Dec 2006 19:43:11 -0000 Subject: Connectionists: new paper on optimal tuning curves Message-ID: <4CCFAF32FB7B2C449A189C82198E426513D32FC8@EXCHVS1.medctr.ad.wfubmc.edu> Dear Colleagues (Apologies for multiple postings) Readers of this list may be interested in this recently published paper. Salinas E (2006) How behavioral constraints may determine optimal sensory representations. PLoS Biol 4(12): e387. DOI: 10.1371/journal.pbio.0040387 ABSTRACT The sensory-triggered activity of a neuron is typically characterized in terms of a tuning curve, which describes the neuron's average response as a function of a parameter that characterizes a physical stimulus. What determines the shapes of tuning curves in a neuronal population? Previous theoretical studies and related experiments suggest that many response characteristics of sensory neurons are optimal for encoding stimulus-related information. This notion, however, does not explain the two general types of tuning profiles that are commonly observed: unimodal and monotonic. Here, I quantify the efficacy of a set of tuning curves according to the possible downstream motor responses that can be constructed from them. Curves that are optimal in this sense may have monotonic or non-monotonic profiles, where the proportion of monotonic curves and the optimal tuning curve width depend on the general properties of the target downstream functions. This dependence explains intriguing features of visual cells that are sensitive to binocular disparity and of neurons tuned to echo delay in bats. The numerical results suggest that optimal sensory tuning curves are shaped not only by stimulus statistics and signal-to-noise properties, but also according to their impact on downstream neural circuits and, ultimately, on behavior. The published version is freely available at: http://biology.plosjournals.org/perlserv/?request=get-document&doi=10.13 71%2Fjournal.pbio.0040387 Cheers, Emilio ----------------------------------------- Emilio Salinas Assistant Professor Department of Neurobiology and Anatomy Wake Forest University School of Medicine Winston-Salem NC 27157 Phone: (336) 713-5176 Fax: (336) 716-4534 e-mail: esalinas at wfubmc.edu www.wfubmc.edu/nba/faculty/salinas ----------------------------------------- From ingber at alumni.caltech.edu Mon Dec 18 16:18:55 2006 From: ingber at alumni.caltech.edu (Lester Ingber) Date: Mon, 18 Dec 2006 21:18:55 -0000 Subject: Connectionists: Paper: Portfolio of Physiological Indicators Message-ID: <20061218211737.GA129@alumnus> The following working paper is available from http://www.ingber.com or http://alumnus.caltech.edu/~ingber %A L. Ingber %T Statistical mechanics of neocortical interactions: Portfolio of Physiological Indicators %R Report 2006:PPI %I Lester Ingber Research %D 2006 %O URL http://www.ingber.com/smni06_ppi.pdf There are several kinds of non-invasive imaging methods that are used to collect data from the brain, e.g., EEG, MEG, PET, SPECT, fMRI, etc. It is difficult to get resolution of information processing using any one of these methods. Approaches to integrate data sources may help to get better resolution of data and better correlations to behavioral phenomena ranging from attention to diagnoses of disease. The approach taken here is to use algorithms developed for the author's Trading in Risk Dimensions (TRD) code using modern methods of copula portfolio risk management, with joint probability distributions derived from the author's model of statistical mechanics of neocortical interactions (SMNI). The author's Adaptive Simulated Annealing (ASA) code is for optimizations of training sets, as well as for importance-sampling. Marginal distributions will be evolved to determine their expected duration and stability using algorithms developed by the author, i.e., PATHTREE and PATHINT codes. Lester From Eugene.Izhikevich at nsi.edu Mon Dec 18 17:23:01 2006 From: Eugene.Izhikevich at nsi.edu (Eugene M. Izhikevich) Date: Mon, 18 Dec 2006 22:23:01 -0000 Subject: Connectionists: Scholarpedia update Message-ID: <45871504.9090400@nsi.edu> Scholarpedia is a free peer-reviewed encyclopedia that combines the philosophies of Wikipedia and Encyclopedia Britannica. Scholarpedia hosts Encyclopedia of Computational Neuroscience, Encyclopedia of Dynamical Systems, and Encyclopedia of Computational Intelligence: http://www.scholarpedia.org/article/Encyclopedia_of_Computational_Neuroscience http://www.scholarpedia.org/article/Encyclopedia_of_Dynamical_Systems http://www.scholarpedia.org/article/Encyclopedia_of_Computational_Intelligence All three will be published in a printed form, and will be used as seeds to start Encyclopedia of Cognitive Neuroscience, Encyclopedia of Applied Mathematics, and Encyclopedia of Computer Science (later next year.) To participate in this project, register with the invitation key 'cns'. Scholarpedia update: Number of registered users: 814 Number of registered curators: 381 Highest Scholar Index: 36 (Ed Ott and Jeff Moehlis) Number of reserved articles: 464 (by curators) Number of finished articles: 71 Number of peer-reviewed and accepted articles: 35 see http://www.scholarpedia.org/article/Special:Allpages Most cited article (53 links): Moehlis J., Josic K., Shea-Brown E.T. (2006) Periodic Orbit. Scholarpedia. p.2464 5 recent articles in Neuroscience: Selverston A. (2006) Stomatogastric Ganglion. Scholarpedia Redgrave P. (2006) Basal Ganglia. Scholarpedia. Chacron M. (2006) Electrolocation. Scholarpedia. Destexhe A. (2006) Spike-and-Wave Oscillations. Scholarpedia Nunez P. and Srinivasan R. (2006) Electroencephalogram. Scholarpedia 5 recent articles in Dynamical Systems: Sanders J.A. (2006) Averaging. Scholarpedia, p.5464 Murdock J. (2006) Unfoldings. Scholarpedia, p.6034 Starmer F. (2006) Initiation of Excitation Waves. Scholarpedia Meiss J. (2006) Dynamical Systems. Scholarpedia Carr J. (2006) Center Manifold. Scholarpedia 5 recent articles in Computational Intelligence: Hutter M. (2006) Algorithmic Information Theory. Scholarpedia. Woergoetter F. and Porr B. (2006) Reinforcement Learning. Scholarpedia Hopfield J.J. (2006) Hopfield Network. Scholarpedia Kohonen T. and Honkela T. (2006) Kohonen Network. Scholarpedia Zadeh L. (2006) Fuzzy Logic. Scholarpedia Number of articles with ongoing election of authors: 62 Please, vote for authors in your area of expertise at http://www.scholarpedia.org/article/Special:ElectionForum To participate in the election, you need to register with the invitation key 'cns'. The invitation key will expire by the end of the year. Eugene M. Izhikevich - editor-in-chief of Scholarpedia. From terry at salk.edu Mon Dec 18 21:01:55 2006 From: terry at salk.edu (Terry Sejnowski) Date: Tue, 19 Dec 2006 02:01:55 -0000 Subject: Connectionists: NEURAL COMPUTATION - January, 2007 In-Reply-To: Message-ID: Neural Computation - Contents - Volume 19, Number 1 - January 1, 2007 LETTERS Mean-Driven and Fluctuation-Driven Persistent Activity in Recurrent Networks Alfonso Renart, Rubén Moreno-Bote, Xiao-Jing Wang, and Néstor Parga Exact Subthreshold Integration with Continuous Spike Times in Discrete Time Neural Network Simulations Abigail Morrison, Sirko Straube, Hans Ekkehard Plesser, and Markus Diesmann The Road to Chaos by Time-Asymmetric Hebbian Learning in Recurrent Neural Networks Colin Molter, Utku Salihoglu and Hugues Bersini Analysis and Design of Echo State Networks Mustafa C. Ozturk, Dongming Xu, and José C. Principe Invariant Global Motion Recognition in the Dorsal Visual System: A Unifying Theory Edmund T. Rolls and Simon M. Stringer Recurrent Cerebellar Loops Simplify Adaptive Control of Redundant and Nonlinear Motor Systems John Porrill and Paul Dean Free-Lunch Learning: Modeling Spontaneous Recovery of Memory J.V. Stone and P.E. Jupp Linear Multilayer ICA Generating Hierarchical Edge Detectors Yoshitatsu Matsuda and Kazunori Yamaguchi Functional Network Topology Learning and Sensitivity Analysis Based on ANOVA Decomposition Enrique Castillo, Noelia Sánchez-Marońo, Amparo Alonso-Betanzos, and Carmen Castillo Second-Order Cone Programming Formulations for Robust Multiclass Classification Ping Zhong and Masao Fukushima Fast Generalized Cross-Validation Algorithm for Sparse Model Learning S. Sundararajan, Shirish Shevade, and S. Sathiya Keerthi ----- ON-LINE - http://neco.mitpress.org/ SUBSCRIPTIONS - 2007 - VOLUME 19 - 12 ISSUES Electronic only USA Canada* Others USA Canada* Student/Retired $60 $63.60 $114 $54 $57.24 Individual $100 $106.00 $154 $90 $95.40 Institution $782 $828.92 $836 $704 $746.24 * includes 6% GST MIT Press Journals, 238 Main Street, Suite 500, Cambridge, MA 02142-9902. Tel: (617) 253-2889 FAX: (617) 577-1545 journals-orders at mit.edu http://mitpressjournals.org/neuralcomp ----- From menno at ics.mq.edu.au Mon Dec 18 22:59:07 2006 From: menno at ics.mq.edu.au (Menno van Zaanen) Date: Tue, 19 Dec 2006 03:59:07 -0000 Subject: Connectionists: Software release: Alignment-Based Learning 1.0 Message-ID: <20061219035850.GN7901@mail.ics.mq.edu.au> ---------------------------------------------------------------------- Software release: ABL 1.0 Alignment-Based Learning ---------------------------------------------------------------------- Alignment-Based Learning (ABL) is a symbolic grammar inference framework that has successfully been applied for several unsupervised machine learning tasks in Natural Language Processing (NLP). Given sequences of symbols only, a system that implements ABL induces structure by aligning and comparing the input sequences. As a result, the input sequences are augmented with the induced structure. More information on the underlying ABL system and its implementation can be found in the following publications: * "Bootstrapping Structure into Language: Alignment-Based Learning", Menno van Zaanen, 2001, PhD Thesis, School of Computing, University of Leeds, UK. http://www.ics.mq.edu.au/~menno/personal_files/docs/t_leeds.ps.gz http://www.ics.mq.edu.au/~menno/personal_files/docs/t_leeds.pdf * "Implementing Alignment-Based Learning", Menno van Zaanen, 2002, In: Proceedings of the International Colloquium on Grammatical Inference (ICGI), pp 312-314, Amsterdam, the Netherlands. http://springerlink.metapress.com/openurl.asp?genre=article&issn=0302-9743&volume=2484&spage=312 * "String alignment in grammatical inference: what suffix trees can do", Jeroen Geertzen, 2003, Technical report ILK-0311, Tilburg University, The Netherlands. http://ilk.uvt.nl/downloads/pub/papers/ilk0311.pdf The package contains a C++ implementation of ABL as well as a reference guide (in texinfo) and the text of the licence. The package should be easy to install on Linux/UNIX systems (using configure). It is maintained by: Menno van Zaanen (menno at ics.mq.edu.au) The latest version of this package can be found on: http://www.ics.mq.edu.au/~menno/research/software/abl/ For any questions, remarks, bugs, improvements, or any other matters of concern about this package, send an email to: menno at ics.mq.edu.au. ---------------------------------------------------------------------- For comments and bug reports relating to ABL, please send mail to: menno at ics.mq.edu.au ---------------------------------------------------------------------- From l.wiskott at biologie.hu-berlin.de Tue Dec 19 07:46:16 2006 From: l.wiskott at biologie.hu-berlin.de (Laurenz Wiskott) Date: Tue, 19 Dec 2006 12:46:16 -0000 Subject: Connectionists: new e-print available about SFA and STDP Message-ID: <17799.57022.263990.437777@huxley.biologie.hu-berlin.de> SLOWNESS: AN OBJECTIVE FOR SPIKE-TIMING-DEPENDENT PLASTICITY? by Henning Sprekeler, Christian Michaelis, and Laurenz Wiskott Slow Feature Analysis (SFA) is an efficient algorithm for learning input-output functions that extract the most slowly varying features from a quickly varying signal. It has been successfully applied to the unsupervised learning of translation-, rotation-, and other invariances in a model of the visual system, to the learning of complex cell receptive fields, and, combined with a sparseness objective, to the self-organized formation of place cells in a model of the hippocampus. In order to arrive at a biologically more plausible implementation of this learning rule, we consider analytically how SFA could be realized in simple linear continuous and spiking model neurons. It turns out that for the continuous model neuron SFA can be implemented by means of a modified version of standard Hebbian learning. In this framework we provide a connection to the trace learning rule for invariance learning. We then show that for Poisson neurons spike-timing-dependent plasticity (STDP) with a specific learning window can learn the same weight distribution as SFA. Surprisingly, we find that the appropriate learning rule reproduces the typical STDP learning window. The shape as well as the timescale are in good agreement with what has been measured experimentally. This offers a completely novel interpretation for the functional role of spike-timing-dependent plasticity in physiological neurons. Available from: http://cogprints.org/5281/ Reference: Sprekeler, H., Michaelis, C., and Wiskott, L. (12. December 2006). Slowness: An Objective for Spike-Timing-Dependent Plasticity? Cognitive Sciences EPrint Archive (CogPrints) 5281, http://cogprints.org/5281/. Also available from: http://itb1.biologie.hu-berlin.de/~wiskott/Abstracts/SpreMichWisk2006.html A short project description is available at: http://itb1.biologie.hu-berlin.de/~wiskott/Projects/SFA-STDP.html -- Prof. Laurenz Wiskott, Institute for Theoretical Biology, Berlin http://itb.biologie.hu-berlin.de/~wiskott/ l.wiskott at biologie.hu-berlin.de From jbednar at inf.ed.ac.uk Tue Dec 19 09:22:37 2006 From: jbednar at inf.ed.ac.uk (James A. Bednar) Date: Tue, 19 Dec 2006 14:22:37 -0000 Subject: Connectionists: PhDs in Neuroinformatics/Computational Neuroscience in Edinburgh Message-ID: <200612191422.kBJEMB4o032433@lodestar.inf.ed.ac.uk> 4 YEAR PhD IN NEUROINFORMATICS AND COMPUTATIONAL NEUROSCIENCE AT THE UNIVERSITY OF EDINBURGH We invite applications for the EPSRC/MRC funded PhD programme in Neuroinformatics/Computational Neuroscience at the University of Edinburgh, UK. This is a 4 year programme with a strongly interdisciplinary character and is ideal for students who want to apply their computational and analytical skills to problems in neuroscience and related fields. The first year consists of courses in neuroscience and informatics, as well as projects based in experimental labs. The first year is followed by a 3 year PhD project. The PhD project is commonly done in collaboration with one of the many departments and institutes affiliated with the DTC. The DTC programme is made up of 3 themes: 1) Computational and Cognitive Neuroscience. Computational, mathematical, and experimental studies of information processing in the nervous system. 2) Neuromorphic Engineering and Robotics. Artificial sensor perception, neuromorphic modelling, spiking computation, and neurorobotics. 3) Data Analysis and Systems. Imaging data analysis and machine learning, Bayesian methods, and building neurally inspired software. Edinburgh has a strong research community in these areas and leads the UK in creating a coherent programme in neuroinformatics. Edinburgh has been voted as 'best place to live in Britain', and has many exciting cultural and student activities. Students with a strong background in computer science, mathematics, physics, or engineering are particularly welcome to apply. Motivated students with other backgrounds will also be considered. About 8 full studentships are available to UK students and a small number of EU students. The stipend is about 12,000 GB pounds per annum. Applicants from outside the EU will need to provide their own funding and evidence thereof. Full info and application forms can be obtained from: http://www.anc.ed.ac.uk/neuroinformatics/. Applications received by March 30th, 2007 will receive priority treatment. From Christian.Igel at neuroinformatik.ruhr-uni-bochum.de Tue Dec 19 17:26:08 2006 From: Christian.Igel at neuroinformatik.ruhr-uni-bochum.de (Christian Igel) Date: Tue, 19 Dec 2006 22:26:08 -0000 Subject: Connectionists: Postdoctoral positions and PhD studentships in Computational Neuroscience Message-ID: <618FED16-B7CF-4B5C-8D35-9B07A476BBFF@neuroinformatik.ruhr-uni-bochum.de> Dear colleagues, We seek motivated postdocs and PhD students for a period of up to 3 years to join our new Bernstein research group "The grounding of higher brain function in dynamic neural fields" at the Ruhr-University Bochum, Germany: http://www.computational-neuroscience-bochum.de The Bernstein group aims at a research program in which theory and experiment are strongly coupled to uncover how higher neural function, including behavior and elementary forms of cognition, emerge from spatio-temporally continuous neural dynamics that are closely coupled to the sensory and motor surfaces and adaptive to sensory and motor experience. The research themes are organized around the concept of dynamic neuronal activation fields, which are estimated from neuronal data and described by neuronal network models to understand perceptual pattern formation, learning, and decision making. The successful candidates will work together with the principal investigators of the research group, Gregor Sch?ner, Dirk Jancke, and Christian Igel. There is a large and lively neuroscience community in Bochum involving the faculties of Biology, Medicine, Psychology, and Chemistry as well as the Institut f?r Neuroinformatik. The Ruhr-University Bochum also hosts an International Graduate School of Neuroscience and a collaborative research center on neural vision. Although we seek individuals with multidisciplinary interests, applicants may be either more biased towards the computational or towards the experimental sciences. Qualified experimentalists will combine novel in vivo voltage-sensitive dye imaging with electrophysiological techniques in visual cortical areas. Candidates for the PhD student positions with a strong mathematical background and a keen interest in neuroscience are particularly welcome. Experience in modeling using dynamical systems is a plus. Potential postdocs should have a strong research record in (computational) neuroscience. Salary is according to the German federal pay scale TVoD 13 and TVoD 13/2 for postdoctoral and PhD student positions, respectively (TVoD 13 corresponds to the former Bat IIa). Applications will be accepted until the positions are filled. Possible starting date is 1 March 2007. The Ruhr-University Bochum is an equal opportunity employer. Inquiries and applications (including CV, statement of interest, names of potential references) should be send via email to info at computational-neuroscience-bochum.de From mhb0 at lehigh.edu Thu Dec 21 09:18:00 2006 From: mhb0 at lehigh.edu (Mark H. Bickhard) Date: Thu, 21 Dec 2006 09:18:00 -0500 Subject: Connectionists: Interactivist Summer Institute 2007: 2nd CFP Message-ID: <88b38d3ac0f89d828bc0a68da49c13ac@lehigh.edu> The web site for the Interactivist Summer Institute 2007, to be held May 26-29, 2007, in Paris, France, is available at: http://www.lehigh.edu/~interact/isi2007/index.htm GENERAL INFORMATION Interactivist Summer Institute 2007 May 26-29, 2007 The American University Paris, France Join us in exploring the frontiers of understanding of life, mind, and cognition. There is a growing recognition - across many disciplines - that phenomena of life and mind, including cognition and representation, are emergents of far-from-equilibrium, interactive, autonomous systems. Mind and biology, mind and agent, are being re-united. The classical treatment of cognition and representation within a formalist framework of encodingist assumptions is widely recognized as a fruitless maze of blind alleys. From neurobiology to robotics, from cognitive science to philosophy of mind and language, dynamic and interactive alternatives are being explored. Dynamic systems approaches and autonomous agent research join in the effort. The interactivist model offers a theoretical approach to matters of life and mind, ranging from evolutionary- and neuro-biology (including the emergence of biological function) through representation, perception, motivation, memory, learning and development, emotions, consciousness, language, rationality, sociality, personality and psychopathology. This work has developed interfaces with studies of central nervous system functioning, the ontology of process, autonomous agents, philosophy of science, and all areas of psychology, philosophy, and cognitive science that address the person. The conference will involve both tutorials addressing central parts and aspects of the interactive model, and papers addressing current work of relevance to this general approach. We will also have a special theme afternoon focusing on Value Systems and the Development of Higher Order Motivation. This will be our fourth Summer Institute; the first was in 2001 at Lehigh University, Bethlehem, PA, USA, the second in 2003 in Copenhagen, Denmark, and the third in 2005 at Clemson University, South Carolina, USA. The Summer Institute is a biennial meeting where those sharing the core ideas of interactivism will meet and discuss their work, try to reconstruct its historical roots, put forward current research in different fields that fits the interactivist framework, and define research topics for prospective graduate students. People working in philosophy of mind, linguistics, social sciences, artificial intelligence, cognitive robotics, theoretical biology, and other fields related to the sciences of mind are invited to send their paper submission or statement of interest for participation to the organizers. Mark H. Bickhard Lehigh University 17 Memorial Drive East Bethlehem, PA 18015 mark at bickhard.name http://bickhard.ws/ From jeedward at gmail.com Fri Dec 22 11:02:55 2006 From: jeedward at gmail.com (John Edward) Date: Fri, 22 Dec 2006 11:02:55 -0500 Subject: Connectionists: Call for papers Message-ID: <000b01c725e2$a58e46f0$6401a8c0@cisnotebookbp> The 2007 International Conference on Artificial Intelligence and Pattern Recognition (AIPR-07) is offering a special session on computational neuroscience. The topics of interest include but will not be limited to * computational neurobiology * realistic simulations of biological neural systems * modeling software and that for computational morphology * phase plane and spike train analysis * web based neuroinformatics. The webpage: http://www.promoteresearch.org contains more details of this conference and other conferences that are being held simultaneously during July 9-12 in Orlando, FL, USA Pease include the statement "computational neuroscience" in the subject line of your email and attach the paper to the email. The email address is: jeedward at gmail.com . John Edward Publicity Committee Co-Chair From mhb0 at lehigh.edu Wed Dec 6 19:50:02 2006 From: mhb0 at lehigh.edu (Mark H. Bickhard) Date: Thu, 07 Dec 2006 00:50:02 -0000 Subject: Connectionists: Interactivist Summer Institute 2007 Message-ID: The web site for the Interactivist Summer Institute 2007, to be held May 26-29, 2007, in Paris, France, is now open: http://www.lehigh.edu/~interact/isi2007/index.htm GENERAL INFORMATION Interactivist Summer Institute 2007 May 26-29, 2007 American University Paris, France Join us in exploring the frontiers of understanding of life, mind, and cognition. There is a growing recognition - across many disciplines - that phenomena of life and mind, including cognition and representation, are emergents of far-from-equilibrium, interactive, autonomous systems. Mind and biology, mind and agent, are being re-united. The classical treatment of cognition and representation within a formalist framework of encodingist assumptions is widely recognized as a fruitless maze of blind alleys. From neurobiology to robotics, from cognitive science to philosophy of mind and language, dynamic and interactive alternatives are being explored. Dynamic systems approaches and autonomous agent research join in the effort. The interactivist model offers a theoretical approach to matters of life and mind, ranging from evolutionary- and neuro-biology (including the emergence of biological function) through representation, perception, motivation, memory, learning and development, emotions, consciousness, language, rationality, sociality, personality and psychopathology. This work has developed interfaces with studies of central nervous system functioning, the ontology of process, autonomous agents, philosophy of science, and all areas of psychology, philosophy, and cognitive science that address the person. The conference will involve both tutorials addressing central parts and aspects of the interactive model, and papers addressing current work of relevance to this general approach. We will also have a special theme afternoon focusing on Value Systems and the Development of Higher Order Motivation. This will be our fourth Summer Institute; the first was in 2001 at Lehigh University, Bethlehem, PA, USA, the second in 2003 in Copenhagen, Denmark, and the third in 2005 at Clemson University, South Carolina, USA. The Summer Institute is a biennial meeting where those sharing the core ideas of interactivism will meet and discuss their work, try to reconstruct its historical roots, put forward current research in different fields that fits the interactivist framework, and define research topics for prospective graduate students. People working in philosophy of mind, linguistics, social sciences, artificial intelligence, cognitive robotics, theoretical biology, and other fields related to the sciences of mind are invited to send their paper submission or statement of interest for participation to the organizers. Mark H. Bickhard Lehigh University 17 Memorial Drive East Bethlehem, PA 18015 mark at bickhard.name http://bickhard.ws/ From cyril.goutte at nrc-cnrc.gc.ca Tue Dec 19 17:10:29 2006 From: cyril.goutte at nrc-cnrc.gc.ca (Cyril Goutte) Date: Tue, 19 Dec 2006 22:10:29 -0000 Subject: Connectionists: CFP: Workshop " Confidence Estimation for Natural Language Processing" at HLT/NAACL 2007 Message-ID: Dear connectionists, The following CFP may be of interest to people working on NLP applications and/or calibration issues. Cyril. -- Cyril.Goutte at nrc-cnrc.gc.ca http://iit-iti.nrc-cnrc.gc.ca/ Interactive Language Technology 101 rue St-Jean-Bosco Institute for Information Technology Gatineau, QC K1A 0R6 National Research Council Canada =================== Confidence Estimation for Natural Language Processing NAACL-HLT 2007 Workshop Rochester, New York, 26 April 2007 http://cenlp.iltevents.org/ =================== Description: Confidence estimation (CE) is a method for automatically estimating the reliability of output generated by NLP systems. It is useful in scenarios where the underlying systems make errors, especially when human users are involved. This is the case in many areas of NLP, because systems are known to be imperfect. Applications can exploit knowledge of likely mistakes to improve their interactions with users and thereby enhance their overall effectiveness. The workshop aims at bringing researchers from different NLP fields together. We would like to provide a forum for sharing ideas, techniques, and experience on common aspects of CE across different areas of NLP. We hope to provide an opportunity for researchers to learn from previous and ongoing work in other fields. =================== Main topics: The workshop invites technical papers related to confidence estimation for natural language processing. Possible topics include, but are not limited to: * CE techniques for different NLP areas * Design and training of CE models; machine learning * Confidence features exploring different knowledge sources * Evaluation and calibration of CE * CE with imprecise or unreliable references * Applications of CE in NLP * Lessons learned from CE systems deployed in applications =================== Submission instructions: Researchers interested in presenting their work at the workshop should prepare a PDF (or postscript) version of their paper, maximum 8 pages (including references) in the NAACL/HLT 2007 format. Reviewing will be double-blind. Submissions will be handled electronically using the START system. The dates are given below. =================== Program Committee: Eugene Agichtein (Emory University, Atlanta) Aron Culotta (U of Massachusetts, Amherst) Chris Drummond (NRC Canada, Ottawa/U of Ottawa) Marcello Federico (ITC-irst, Trento) Simona Gandrabur (Idilia, Montreal) Christian Gollan (RWTH Aachen University) Didier Guillevic (Idilia, Montreal) Philippe Langlais (RALI, U de Montreal) Hermann Ney (RWTH Aachen University) Chris Quirk (Microsoft Research, Redmond) Carl Edward Rasmussen (Max Planck Institute, T?bingen) Alberto Sanchis (U of Valencia) Michel Simard (NRC Canada, Gatineau) Enrique Vidal (U of Valencia) Stephan Vogel (CMU, Pittsburgh) =================== Organizers: Nicola Ueffing (NRC Canada, Gatineau) George Foster (NRC Canada, Gatineau) Cyril Goutte (NRC Canada, Gatineau) =================== Important Dates: Submission deadline: 18 Jan 2007 Notification: 22 Feb 2007 Camera-ready: 1 Mar 2007 Workshop date: 26 Apr 2007 For more information, see the workshop page at http://cenlp.iltevents.org/ --- Nicola Ueffing Interactive Language Technologies Group National Research Council Canada 101 St-Jean-Bosco Street Gatineau, QC, Canada Tel.: +1 (819) 934-8992 Fax: +1 (819) 934-2607 Email: Nicola.Ueffing at cnrc-nrc.gc.ca From Johan.Suykens at esat.kuleuven.be Fri Dec 22 09:46:29 2006 From: Johan.Suykens at esat.kuleuven.be (Johan Suykens) Date: Fri, 22 Dec 2006 15:46:29 +0100 Subject: Connectionists: International Symposium on Synchronization in Complex Networks (SynCoNet2007) Message-ID: <458BEFC5.3070207@esat.kuleuven.ac.be> (apologies for multiple postings, please feel free to further distribute) ============================================================================= International Symposium on Synchronization in Complex Networks (SynCoNet2007) Leuven Belgium, July 2-4, 2007 http://www.esat.kuleuven.ac.be/sista/synconet2007/ ============================================================================= * Scope Nowadays the main focus in the study of synchronization phenomena is moving from the analysis of a low number of interacting units to systems consisting of thousands and even millions of nodes. This is relevant for the study of problems in genetic networks, neural networks, networks in social sciences, power grid networks, coupled mechanical systems, information networks and others. It poses new challenges for the understanding of cluster formation, stability, sensitivity and robustness, bifurcations and chaos etc. The scope of the SynCoNet 2007 symposium is to provide a multi-disciplinary forum where researchers can meet in this fascinating area at the intersection between synchronization and pattern formation in complex networks. * Invited Speakers Igor Belykh, Georgia State University Guanrong Chen, City University of Hong Kong Alexander Fradkov, Russian Academy of Sciences St. Petersburg Martin Hasler, Ecole Polytechnique Federale de Lausanne Juergen Kurths, University of Potsdam Erik Mosekilde, Technical University of Denmark Henk Nijmeijer, Eindhoven University of Technology Edward Ott, University of Maryland Louis Pecora, Naval Research Laboratory Washington DC * Format The SynCoNet 2007 program will feature invited talks together with oral sessions and poster sessions. Interested participants are cordially invited to submit an extended abstract (max. 2 pages) for a symposium contribution. The journal CHAOS will publish a focus issue on Synchronization in Complex Networks (tentatively scheduled Sept. 2008). This focus issue will consist of selected papers (invited) from the SynCoNet 2007 symposium. * Schedule Extended abstract submission: Feb. 20, 2007 Notification of acceptance: April 1, 2007 Final version of extended abstract: May 7, 2007 Deadline for registration: June 1, 2007 Symposium: July 2-4, 2007 * Organizing Committee Co-chairs: Grigory Osipov, Nizhny Novgorod University Johan Suykens, K.U.Leuven Dirk Aeyels, UGent Vladimir Belykh, Nizhny Novgorod University Bart Motmans, K.U.Leuven Joos Vandewalle, K.U.Leuven * Location and venue The symposium will take place in Leuven (Heverlee) Belgium in the Arenberg Castle. The place is surrounded by several restaurants/cafes and parks where one may have a relaxing walk. The historical town of Leuven is within walking distance from the meeting site. Leuven is also well-known for its pleasant atmosphere, pubs and restaurants. * More information: http://www.esat.kuleuven.ac.be/sista/synconet2007/ Disclaimer: http://www.kuleuven.be/cwis/email_disclaimer.htm From malchiodi at dsi.unimi.it Fri Dec 22 10:41:49 2006 From: malchiodi at dsi.unimi.it (Dario Malchiodi) Date: Fri, 22 Dec 2006 16:41:49 +0100 Subject: Connectionists: (no subject) Message-ID: <017B4B8F-F939-47EC-8B17-9D6D37B4A24C@dsi.unimi.it> Apologizes for cross-posting. ------------------------------------ CFP: Invited session on Learning from uncertain data KES2007: 11th International Conference on Knowledge-Based and Intelligent Information & Engineering Systems in conjunction with WIRN07: 17th Italian Workshop on Neural Networks September 12-14, 2007 - Vietri sul Mare (SA), Italy http://homes.dsi.unimi.it/~malchiod/WIRN07/ ------------------------------------ The session will focus on research activities in the field of learning from data when a quantitative measure of the uncertainty affecting each data item is available, as a counterpart of the usual samples in input to learning algorithms, gathering data to be considered homogeneously. ------------------------------------ Topics The session will welcome (although not limiting to) contributions on theoretical analysis of the problem, design of machine learning algorithms, as well as real-world applications. A non exhaustive list of topics to focus on is the following: * Learning algorithms * Integration between learning and fuzzy sets * Soft computing methods for classification and model selection * Support vector machines and kernel methods * Integration between optimization and learning * Statistical methods for dealing with uncertainty * Real-world applications of uncertainty-based learning algorithms ------------------------------------ Paper submissions Participants are encouraged to submit papers (up to eight pages in length) in the Springer LNCS/LNAI format. The conference proceedings will be published by Springer-Verlag in the LNAI series. All submissions should be sent in PDF to malchiodi(AT)dsi.unimi.it. ------------------------------------ Important dates Paper deadline: February 28th, 2007 Notification of acceptance: April 1st, 2007 Camera-ready papers: April 30th, 2007 ------------------------------------ Session chair Dario Malchiodi Dipartimento di Scienze dell'Informazione Universit? degli Studi di Milano Via Comelico 39/41 20135 Milano ITALY T: +39 02 503 16338 F: +39 02 503 16276 E: malchiodi(AT)dsi.unimi.it W: http://homes.dsi.unimi.it/~malchiod ------------------------------------ Conference venue The conference will be organized in the beautiful town of Vietri sul Mare, near Salerno. Please refer to the official KES2007 web site for details on the conference and for travel informations. From juergen at idsia.ch Mon Dec 25 10:44:04 2006 From: juergen at idsia.ch (Juergen Schmidhuber) Date: Mon, 25 Dec 2006 16:44:04 +0100 Subject: Connectionists: Goedel Machines, Universal AI, Consciousness Message-ID: <0162f78624d28c568f5fd6613cfbe6f3@idsia.ch> Kurt Goedel was born 100 years ago. 75 years ago he layed the foundations of theoretical computer science with his work on universal formal languages and the limits of proof and computation (1931), just 5 years after Julius Lilienfeld patented the transistor, and 10 years before Konrad Zuse completed the first working program-controlled computer (1941). Goedel (or G?del) constructed self-referential statements that talk about themselves, in particular, about whether they can be derived from a set of given axioms through a computational theorem proving procedure. He went on to construct statements that deny their own provability, to demonstrate that traditional math is either flawed in a certain sense or contains unprovable but true statements. We use Goedel's self-reference trick to build a universal, fully self-referential, self-improving, optimally efficient problem solver. A Goedel Machine is a computer whose original software includes axioms describing the hardware and the original software (this is possible without circularity) plus whatever is known about the (probabilistic) environment plus some formal goal in form of an arbitrary user-defined utility function, e.g., cumulative future expected reward in a sequence of optimization tasks. The original software also includes a proof searcher which uses the axioms (and possibly an online variant of Levin's universal search) to systematically make pairs ("proof'', "program'') until it finds a proof that a rewrite of the original software through "program'' will increase utility. The machine can be designed such that each self-rewrite is necessarily globally optimal in the sense of the utility function, even those rewrites that destroy the proof searcher. If we equate the notion of consciousness with the ability to execute unlimited formal self-inspection and provably useful self-change (unlimited except for the limits of computability and provability), then the GM and its Global Optimality Theorem can be viewed as providing the first technical justification of consciousness. Goedel Machine papers up to 2006 (Goedel Centenary): http://www.idsia.ch/~juergen/goedelmachine.html Juergen Schmidhuber From takane at takane2.psych.mcgill.ca Thu Dec 28 11:58:04 2006 From: takane at takane2.psych.mcgill.ca (YoshioTakane) Date: Thu, 28 Dec 2006 11:58:04 -0500 Subject: Connectionists: IMPS-2007 Message-ID: <000d01c72aa1$57878930$3449fea9@psych.mcgill.ca> CALL For PAPERS IMPS (International Meeting of the Psychometric Society) - 2007 will be held in Tokyo between July 9 (Monday) - 13 (Friday), 2007. I am organizing a symposium on Neural Network (NN) Models at the meeting, and would like to solicit your participation as an oral presenter in the symposium. Topics include: 1) Nonlinear information integration, 2) Metahypothesis and prior knowledge in various forms, 3) Knowledge transformation and transfer, 4) Knowledge representation, etc. Detailed information about the meeting can be found at: http://www.psychometricsociety.org/ Please submit the title and the abstract (200 words or less) of your presentation to takane at psych.mcgill.ca preferrably by January 15 but no later than January 31, 2007. Yoshio Takane Professor of Psycholgy McGill University