simulation@uflorida.cis.ufl.edu (Moderator: Paul Fishwick) (08/10/89)
Volume: 10, Issue: 4, Wed Aug 9 17:30:41 EDT 1989 +----------------+ | TODAY'S TOPICS | +----------------+ (1) TR available: NN's in Simulation (2) Object-Oriented Simulation/Positions Available (3) REAL Network Simulator Available for FTPing * Moderator: Paul Fishwick, Univ. of Florida * Send topical mail to: simulation@uflorida.cis.ufl.edu * Archives available via FTP to bikini.cis.ufl.edu, login as 'anonymous', use your last name as the password, change directory to pub/simdigest. * Simulation Tools available by doing above and changing the directory to pub/simdigest/tools. ----------------------------------------------------------------------------- To: comp-simulation@ufl.edu Path: uflorida!fish.cis.ufl.edu!fishwick From: fishwick@fish.cis.ufl.edu (Paul Fishwick) Newsgroups: comp.ai.neural-nets,comp.simulation Subject: neural nets in system identification Date: 7 Aug 89 21:26:46 GMT Sender: news@bikini.cis.ufl.edu Reply-To: fishwick@fish.cis.ufl.edu () Organization: UF CIS Department [[They seem to be putting out TR announcements in other news groups, and this seems like a good way of keeping news items flowing - please send any simulation-related dissertation/thesis/tech.report info that you have --- we need a good way of knowing about each other's work in the "pre-print" stage. -PAF ]] For a pre-print, please e-mail to paulette@bikini.cis.ufl.edu and request tech report # (TR-89-8). NEURAL NETWORK MODELS IN SIMULATION: A COMPARISON WITH TRADITIONAL MODELING APPROACHES Paul A. Fishwick Department of Computer and Information Science University of Florida Bldg. CSE, Room 301 Gainesville, FL 32611 To be presented at: The Winter Simulation Conference, Dec. 1989 ABSTRACT Neural models are enjoying a resurgence in systems research primarily due to a general interest in the connectionist approach to modeling in artificial intelligence and to the availability of faster and cheaper hardware on which neural net simulations can be executed. We have experimented with using a multi-layer neural network model as a simulation model for a basic ballistics model. In an effort to evaluate the efficiency of the neural net implementation for simulation modeling, we have compared its performance with traditional methods for geometric data fitting such as linear regression and surface response methods. Both of the latter approaches are standard features in many statistical software packages. We have found that the neural net model appears to be inadequate in most respects and we hypothesize that accuracy problems arise, primarily, because the neural network model does not capture the system structure characteristic of all physical models. We discuss the experimental procedure, issues and problems, and finally consider possible future research directions. +------------------------------------------------------------------------+ | Prof. Paul A. Fishwick.... INTERNET: fishwick@bikini.cis.ufl.edu | | Dept. of Computer Science. UUCP: gatech!uflorida!fishwick | | Univ. of Florida.......... PHONE: (904)-335-8036 | | Bldg. CSE, Room 301....... FAX is available | | Gainesville, FL 32611..... | +------------------------------------------------------------------------+ ------------------------------ Date: Tue, 8 Aug 89 09:48:16 EST From: Mark D. Grover <grover@potomac.ads.com> To: simulation@bikini.cis.ufl.edu Subject: Obj-oriented Simulation/Graphics/AI (Symbolics,Silicon Graphics) I have just posted a description of new job opportunities in land wargaming and in cockpit simulation creation to misc.jobs.offered. If you do not receive this newsgroup, feel free to send mail to me for a copy of the announcement. These positions are for our northern Virginia research office, although we work closely with our main office in Mountain View, California. They involve design and creation of simulations, novel simulation techniques, and the use of Symbolics and Silicon Graphics computers. Mark D. Grover (grover@Potomac.ADS.COM) 703-243-1611 Advanced Decision Systems 1500 Wilson Blvd #512; Arlington, VA 22209 There are terrible people who, instead of solving a problem, bungle it and make it more difficult for all who come after. Whoever can't hit the nail on the head should, please, not hit it at all. --- Friedrich Nietzsche ------------------------------ Date: Wed, 9 Aug 89 12:24:29 -0700 From: keshav@ucbarpa.Berkeley.EDU (Srinivasan Keshav) To: fishwick@bikini.cis.ufl.edu Subject: Submission for comp.simulation The REAL network simulator, based on the NEST simulation package from Columbia, is ready for anonymous ftp from ucbarpa.berkeley.edu. The tar format files are in REAL/archive, and REAL/README tells you how to unpack it. REAL will run on Sun 3s, Vaxen and the Mips machine under 4.3BSD. It does realistic and fast simulation of transport layer protocols with specific reference to congestion control. REAL is fully documented, and the documentation is part of the release. Further documentation on NEST is available from dupuy@cs.columbia.edu. User licencing is free for non-commercial use. Licencing procedures and forms are included in the package and are described in the documentation. Direct comments or questions to me. keshav (keshav@ucbarpa.Berkeley.EDU) ------------------------------ END OF SIMULATION DIGEST ************************