neuron-request@HPLMS2.HPL.HP.COM ("Neuron-Digest Moderator Peter Marvit") (01/13/91)
Neuron Digest Saturday, 12 Jan 1991 Volume 7 : Issue 4 Today's Topics: Regarding Brain Size and Sulci neural nets for fingerprint recognition Applications summary. Summary of medical/astronomical imaging w/ANNs Connectionist Simulators Transputers for neural networks? Kohonen's Network again Call for Papers: Uncertainty in AI 91 Machine learning workshop: Addendum to call for papers Call for papers Send submissions, questions, address maintenance and requests for old issues to "neuron-request@hplabs.hp.com" or "{any backbone,uunet}!hplabs!neuron-request" Use "ftp" to get old issues from hplpm.hpl.hp.com (15.255.176.205). ------------------------------------------------------------ Subject: Regarding Brain Size and Sulci From: UAP001%DDOHRZ11.BITNET@CUNYVM.CUNY.EDU Date: Wed, 19 Dec 90 10:28:56 +0700 On the matter of brain size and number of sulci: exceptions -e.g., Gauss- have been reported; but there's no clear pattern. A century ago measuring brains of famous savants was popular, but went out of fashion when it turned out that geniuses tended to fall into the normal range. Incidentally, although we don't know the surface area of their cortex, the Neanderthal probably had slightly larger brains than ours. ******************************************************************** * C.R. Cavonius BITNET:uap001@ddohrz11 * * Inst. f. Arbeitsphysiologie (Note: uap-zero-zero-one, * * an der Universitaet Dortmund not uap-oh-oh-one) * * Ardeystr. 67 Tel: +49 231 1084 261 * * D-4600 Dortmund 1, F.R. Germany Fax: +49 231 1084 308 * ******************************************************************** ------------------------------ Subject: neural nets for fingerprint recognition From: oflazer%TRBILUN.BITNET@CUNYVM.CUNY.EDU Date: Mon, 24 Dec 90 18:25:00 +0200 Hello, I would appreciate any pointers to work on recognition of finger prints using a neural network approach. Thanks Kemal Oflazer oflazer@trbilun.bitnet (BITNET) oflazer%trbilun.bitnet@cunyvm.cuny.edu (INTERNET) ------------------------------ Subject: Applications summary. From: yuhas@faline.bellcore.com (Ben Yuhas) Date: Wed, 26 Dec 90 15:16:02 -0500 I am trying to gather a list of neural networks applications that have, or are about to, become commercial products. At NIPS we heard about such work from FORD and Sarnoff labs, I would appreciate any other examples that any of you are aware of. Ben yuhas@bellcore. ------------------------------ Subject: Summary of medical/astronomical imaging w/ANNs From: Denis Anthony <esrmm@cu.warwick.ac.uk> Organization: Computing Services, Warwick University, UK Date: Sat, 29 Dec 90 15:31:51 +0000 [[ Editor's Note: My thanks to Denis for providing this summary. My frequent theme is for readers who request information to share the resulst of their search with the Digest so all may benefit. Readers who post what they're doing may find kindred spirits in unlikely places as well! -PM ]] Peter You stated in the Digest that you would like a summary of the search in medical/astronomical imaging using neural nets. I give below short notes on the small number of replies I have had, and a short summary of what I am doing. I am not sure how much use the below is, but hopefully the names of interested users may be of some interest. nb. I only have medical replies here. Denis 1. Peter J.B. Hancock (pjh@cs.stir.ac.uk) Starting looking at MRI brain data, with a view to describing the lesions caused by accidents, and eventually trying to predict the long-term recovery of the patient. 2. Paul Marquis (pmarquis@jade.tufts.edu (Internet) PTMARQUI@TUFTS (BITNET)) I'm a graduate student at Tufts University and am currently considering using neural networks as an approach to Magnetic Resonance. 3. Henk D. Davids (hd%nl.philips.cft.philtis%uucp.phigate@nl.nluug.hp4nl) We are working with medical images too. I would love to see more application of nn technology to medical image data discussed here. Seems to me that there are some problems that are specific for this kind of processing: the large input space ... I think that there must be sufficient interest to warrant discussion on Usenet rather than taking this type of application into a private corner. 4. Nathan Combs. (ncombs%fandago.ctc.tasc.com%fandago@net.uu.uunet) Steve Stone (STONES@edu.wsu.csc.wsuvm1) Rajiv Sarathy ( INTERNET sarathy@gpu.utcs.utoronto.ca) Above expressed interest in ANNs and medical imaging. 5. Denis Anthony (esrmm@cu.warwick.ac.uk) Previously working on classification of lung scintigrams using neural networks. Now working on the inverse problem of ultrasound tomography using nns. References below may be emailed to any interested parties. %Q Anthony D.M, Hines E.L, Taylor D, and Barham J %D 1989 %T A Study of Data Compression using Neural Networks and Principal Component Analysis (IEE Coloquiuum on Biomedical Applications of Digital Signal Processing) %P 2/1-2/4 %Q Anthony D.M, Hines E.L, Taylor D, and Barham J %D 1990 %T The Use of Genetic Algorithms to Learn the Most Appropriate Inputs to a Neural Network %J IASTED Conf. Artificial Intelligence App. Neural Networks %Q Anthony D.M, Hines E.L, Taylor D, and Barham J %D 1990 %T The Use of Neural Networks to Classify Lung Scintigrams %J IASTED Conference on Applied Informatics %P 240-242 %Q Anthony D.M, Hines E.L, Taylor D, and Barham J %D 1989 %T An Investigation into the Use of Neural Networks for an Expert System in Nuclear Medicine Image Analysis %J IEE Conference on Image Processing %P 338-342 %Q Anthony D.M, Hines E.L, Taylor D, and Barham J %D 1990 %T The Use of Neural Networks in Classifying Lung Scintigrams %J INNS Int. Neural Networks Conference %V 1 %P 71-74 %Q Anthony D.M %D 1990 %T Reducing Connectivity in Compression Networks %J Neural Network Review %Q Chiu W.C, Anthony D.M, Hines E.L, Forno C, Hunt R, and Oldfield S %D 1990 %T Selection of the Optimal MLP Net for Photogrammetric Target Processing %J IASTED Conf. Artificial Intelligence App. Neural Networks ------------------------------ Subject: Connectionist Simulators From: Kim Daugherty <kimd@gizmo.usc.edu> Date: Mon, 07 Jan 91 17:12:13 -0800 [[ Editor's Note: Once again, I'm reminded of the need of a "canonical" list of neural network simulators. I've been meaning to assemble same, but doubt that I'll have enough free time to do so. Besides this below, there's also the PDP Volume 3 code, Mactivation, and many other public domain or low-cost programs. In addition, there is a host of commercial products. Does anyone keep such a "complete list" which they would share with the Digest? -PM ]] Last November, I posted a request for connectionist modeling simulators to the mailing list. I would like to thank those who responded. Following is a list and brief description of several simulators: 1. Genesis - An elaborate X windows simulator that is particularly well suited for modeling biological neural networks. unix> telnet genesis.cns.caltech.edu (or 131.215.135.185) Name: genesis Follow directions there to get a ftp account from which you can ftp 'genesis.tar.Z". This contains genesis source and several tutorial demos. NOTE: There is a fee to become a registered user. 2. PlaNet (AKA SunNet) - A popular connectionist simulator with versions to run under SunTools, X Windows, and non-graphics terminals created by Yoshiro Miyata. The SunTools version is not supported. unix> ftp boulder.colorado.edu (128.138.240.1) Name: anonymous Password: ident ftp> cd pub ftp> binary ftp> get PlaNet5.6.tar.Z ftp> quit unix> zcat PlaNet5.6.tar.Z | tar xf - All you need to do to try it is to type: unix> Tutorial This will install a program appropriate for your environment and start an on-line tutorial. If you don't need a tutorial, just type 'Install' to install the system and then 'source RunNet' to start it. See the file README for more details. The 60-page User's Guide has been split into three separate postscript files so that each can be printed from a printer with limited memory. Print the files doc/PlaNet_doc{1,2,3}.ps from your postscript printer. See the doc/README file for printing the Reference Manual. Enjoy!! And send any questions to miyata@boulder.colorado.edu. 3. CMU Connectionist Archive - There is a lisp backprop simulator in the connectionist archive. unix> ftp b.gp.cs.cmu.edu (or 128.2.242.8) Name: ftpguest Password: cmunix ftp> cd connectionists/archives ftp> get backprop.lisp ftp> quit 4. Cascade Correlation Simulator - There is a LISP and C version of the simulator based on Scott Fahlman's Cascade Correlation algorithm, who also created the LISP version. The C version was created by Scott Crowder. unix> ftp pt.cs.cmu.edu (or 128.2.254.155) Name: anonymous Password: (none) ftp> cd /afs/cs/project/connect/code ftp> get cascor1.lisp ftp> get cascor1.c ftp> quit A technical report descibing the Cascade Correlation algorithm may be obtained as follows: unix> ftp cheops.cis.ohio-state.edu (or 128.146.8.62) Name: anonymous Password: neuron ftp> cd pub/neuroprose ftp> binary ftp> get fahlman.cascor-tr.ps.Z ftp> quit unix> uncompress fahlman.cascor-tr.ps.Z unix> lpr fahlman.cascor-tr.ps 5. Quickprop - A variation of the back-propagation algorithm developed by Scott Fahlman. A LISP and C version can be obtained in the same directory as the cascade correlation simulator above. Kim Daugherty kimd@gizmo.usc.edu ------------------------------ Subject: Transputers for neural networks? From: "Tom Tollenaere " <ORBAN%BLEKUL13.BITNET@CUNYVM.CUNY.EDU> Date: Wed, 09 Jan 91 12:25:45 +0000 [[ Editor's Note: As usual, I certainly hope that Tom summarizes the results he gets and posts them to the Digest. Of course, readers with relevant information are free to post directly to the Digest (neuron-request@hplabs.hpl.hp.com) as well as sending a copy to Tom. -PM]] Hi netters, I am compiling a technical report on scientific/industrial use of transputers and similar parallel computers (Intel IPSC, NCube, Cosmic Cube, etc...) for neural network research/applications. If you happen to be involved both in neural networks and parallel computers, please get in touch with me. I'm interested in * what machine do you use ? And in what language do you program it ? * what kind of networks do you run on the machine ? feed forward things like perceptrons, backprop, or more dynamic things like Hopfield, Kohonen networks ? * how did you do it ? i.e. how did you handle parallelism * are you happy ? i.e. how well does the application run ? Do you have any performance/efficiency measures ? * what do you use it for ? i.e. do you run simulations for basic network research ? Are you involved with research on a particular application and do you use the parallel machine just c'se its fast ? Do you have an industrial application ? If so, do you sell it? What price ? Happy clients ? The idea is to get an overview of what people are doing with neural nets and parallel computers, how they are doing it, and how well it goes. When the report is finished, I'll post a note on the net, so anyone interested can get a copy. I'm looking forward to massive response; especially from the U.S. and from industrial users. Don't use the bulletin board, but contact me directly. Cheers, tom Tom TOLLENAERE Laboratorium voor Neuro en Psychofysiologie Katholieke Universiteit Leuven Campus Gasthuisberg Herestraat 49 B-3000 Leuven BELGIUM - EUROPE email : ORBAN at blekul13.bitnet or blekul13.earn fax : 32 16 21 59 93 phone : 32 16 21 59 60 Acknowledge-To: <ORBAN@BLEKUL13> ------------------------------ Subject: Kohonen's Network again From: JJ Merelo <jmerelo@ugr.es> Date: 10 Jan 91 12:52:00 +0200 I am working on Kohonen network, and I have met lots of trouble when I have tried to find the correct parameters k1 and k2 on the learning algorithm. Does anyone know how to find them? Besides, the convergence of the learning procedure is guaranteed because of the decreasing nature of the alpha gain factor. But is it guaranteed that it will converge to the right vectors. In other clustering algorithms, it does not end until convergence in clustering mean vectors is reached ( v.g. k-means), and I think this is more correct. By the way, is anybody working on Kohonen's network? I have seen it quoted thousands of times, but the quotes are always from the same papers from Kohonen himself. I know not about anybody who has got Kohonen net working ( maybe Aleksander, as he says in his book, but this is the only one ). I think it *must* work, but I have got mixed results. Besides, it is boring to keep on trying new parameters. I hope to get some help, JJ Merelo Dpto. de Electronica y Sistemas Informaticos Facultad de Ciencias Campus Fuentenueva, s/n 18071 Granada ( Spain ) e-mail JMERELO@UGR.ES ------------------------------ Subject: Call for Papers: Uncertainty in AI 91 From: dambrosi@kowa.CS.ORST.EDU Date: Fri, 21 Dec 90 12:33:02 +0000 THE SEVENTH CONFERENCE ON UNCERTAINTY IN ARTIFICIAL INTELLIGENCE UCLA, Los Angeles July 13-15, 1991 (Preceding AAAI) The seventh annual Conference on Uncertainty in AI is concerned with the full gamut of approaches to automated and interactive reasoning and decision making under uncertainty including both quantitative and qualitative methods. We invite original contributions on fundamental theoretical issues, on the development of software tool embedding approximate reasoning theories, and on the validation of such theories and technologies on challenging applications. Topics of particular interest include: - Foundations of uncertainty - Semantics of qualitative and quantitative uncertainty representations - The role of uncertainty in automated systems - Control of reasoning; planning under uncertainty - Comparison and integration of qualitative and quantitative schemes - Knowledge engineering tools and techniques for building approximate reasoning systems - User Interface: explanation and summarization of uncertain information - Applications of approximate reasoning techniques Papers will be carefully refereed. All accepted papers will be included in the proceedings, which will be available at the conference. Papers may be accepted for presentation in plenary sessions or poster sessions. Five copies of each paper should be sent to the Program Chair by March 4, 1991. Acceptance will be sent by April 22, 1991. Final camera-ready papers, incorporating reviewers' comments, will be due by May 10, 1991. There will be an eight page limit on the camera-ready copy (with a few extra pages available for a nominal fee.) Program Co-Chair: Bruce D'Ambrosio Philippe Smets Dept. of Computer Science IRIDIA 303 Dearborn Hall Universite Libre de Bruxelles. Oregon State University 50 av. Roosevelt, CP 194-6 Corvallis, OR 97331-3202 USA 1050 Brussels, Belgium tel: 503-737-5563 tel: +322.642.27.29 fax: 503-737-3014 fax: +322.642.27.15 e-mail: dambrosio@CS.ORST.EDU e-mail: R01501@BBRBFU01.BITNET General Chair: Piero Bonissone General Electric Corporate Research and Development 1 River Rd., Bldg. K1-5C32a, 4 Schenectady, NY 12301 tel: 518-387-5155 fax: 518-387-6845 e-mail: bonisson@crd.ge.com Program Committee: Piero Bonissone, Peter Cheeseman, Max Henrion, Henry Kyburg, John Lemmer, Tod Levitt, Ramesh Patil, Judea Pearl, Enrique Ruspini, Ross Shachter, Glenn Shafer, Lofti Zadeh. ------------------------------ Subject: Machine learning workshop: Addendum to call for papers From: Lawrence Birnbaum <birnbaum@fido.ils.nwu.edu> Date: Tue, 08 Jan 91 14:54:23 -0600 ADDENDUM TO CALL FOR PAPERS EIGHTH INTERNATIONAL WORKSHOP ON MACHINE LEARNING NORTHWESTERN UNIVERSITY EVANSTON, ILLINOIS JUNE 27-29, 1991 We wish to clarify the position of ML91 with respect to the issue of multiple publication. In accordance with the consensus expressed at the business meeting at ML90 in Austin, ML91 is considered by its organizers to be a specialized workshop, and thus papers published in its proceedings may overlap substantially with papers published elsewhere, for instance IJCAI or AAAI. The sole exception is with regard to publication in future Machine Learning Conferences. Authors who are concerned by this constraint will be given the option of foregoing publication of their presentation in the ML91 Proceedings. The call for papers contained information concerning seven of the eight individual workshops that will make up ML91. Information concerning the final workshop follows. Larry Birnbaum Gregg Collins Northwestern University The Institute for the Learning Sciences 1890 Maple Avenue Evanston, IL 60201 (708) 491-3500 - ------------------------------------------------------------------------------- COMPUTATIONAL MODELS OF HUMAN LEARNING This workshop will foster interaction between researchers concerned with psychological models of learning and those concerned with learning systems developed from a machine learning perspective. We see several ways in which simulations intended to model human learning and algorithms intended to optimize machine learning may be mutually relevant. For example, the way humans learn and the optimal method may turn out to be the same for some tasks. On the other hand, the relation may be more indirect: modeling human behavior may provide task definitions or constraints that are helpful in developing machine learning algorithms; or machine learning algorithms designed for efficiency may mimic human behavior in interesting ways. We invite papers that report on learning algorithms that model or are motivated by learning in humans or animals. We encourage submissions that address any of a variety of learning tasks, including category learning, skill acquisition, learning to plan, and analogical reasoning. In addition, we hope to draw work from a variety of theoretical approaches to learning, including explanation-based learning, empirical learning, connectionist approaches, and genetic algorithims. In all cases, authors should explicitly identify 1) in what ways the system's behavior models human (or animal) behavior, 2) what principles in the algorithm are responsible for this, and 3) the methods for comparing the system's behavior to human behavior and for evaluating the algorithm. A variety of methods have been proposed for computational psychological models; we hope the workshop will lead to a clearer understanding of their relative merits. Progress reports on research projects still in development are appropriate to submit, although more weight will be given to projects that have been implemented and evaluated. Integrative papers providing an analysis of multiple systems or several key issues are also invited. WORKSHOP COMMITTEE Dorrit Billman (Georgia Tech) Randolph Jones (Univ. of Pittsburgh) Michael Pazzani (Univ. of California, Irvine) Jordan Pollack (Ohio State Univ.) Paul Rosenbloom (USC/ISI) Jeff Shrager (Xerox PARC) Richard Sutton (GTE) SUBMISSION DETAILS Papers should be approximately 4000 words in length. Authors should submit seven copies, by March 1, 1991, to: Dorrit Billman School of Psychology Georgia Institute of Technology Atlanta, GA 30332 phone (404) 894-2349 Formats and deadlines for camera-ready copy will be communicated upon acceptance. ------------------------------ Subject: Call for papers From: RAKESH@IBM.COM Date: Wed, 09 Jan 91 11:34:17 -0500 CALL FOR PAPERS Progress In Neural Networks Special Volume on Neural Networks In Vision Significant progress has been made recently in the application of neural networks to computational vision. To showcase this research, Ablex Publishing is planning a special volume on "Neural Networks in Vision", scheduled for 1992. This volume will be a part of "Progress in Neural Networks", an annual book series reviewing research in modelling, analysis, design and application of neural networks. Authors are invited to submit original manuscripts detailing recent progress in neural networks for vision. The paper should be tutorial in nature, self contained and preferably, but not necessarily, about fifty double spaced pages in length. An abstract and an outline are due by January 31, 1991, the full paper by Feburary 28, 1991. Make submissions to Rakesh Mohan Associate Volume Editor IBM Thomas J. Watson Research Center PO Box 704 Yorktown Heights, NY 10598 email: rakesh@ibm.com ------------------------------ End of Neuron Digest [Volume 7 Issue 4] ***************************************