NAME

snns - neural network simulation package SNNS

SYNOPSIS

xgui [options]

batchman [options]

snns2c [options]

DESCRIPTION

SNNS (Stuttgart Neural Network Simulator) is a simulator for neural networks developed at the Institute for Parallel and Distributed High Performance Systems, at the University of Stuttgart since 1989. The goal of the project is to create an efficient and flexible simulation environment for research on and application of neural nets.

The SNNS simulator consists of four main components: Simulator kernel, graphical user interface, batch execution interface batchman(1), and network compiler snns2c(1). The simulator kernel operates on the internal network data structures of the neural nets and performs all operations on them. The graphical user interface xgui(1), built on top of the kernel, gives a graphical representation of the neural networks and controls the kernel during the simulation run. In addition, the user interface can be used to directly create, manipulate and visualize neural nets in various ways. Complex networks can be created quickly and easily. Nevertheless, XGUI should also be well suited for unexperienced users, who want to learn about connectionist models with the help of the simulator. An online help system, partly context-sensitive, is integrated, which can offer assistance with problems.

An important design concept was to enable the user to select only those aspects of the visual representation of the net in which he is interested. This includes depicting several aspects and parts of the network with multiple windows as well as suppressing unwanted information.

COMPATIBILITY

Machines and operating systems on which SNNS has been tested

SUN SparcStation SunOS 4.1.2 - 5.2.6 DECstation Ultrix V4.2 DEC Alpha OSF1 V2.1 - V4.0 IBM-PC 80486, Pentium Linux, NeXTStep IBM RS 6000 AIX V3.1 - V4.1 HP 9000 HP-UX 8.07 - 10.20, NeXTStep SGI Indigo 2 IRIX 4.0.5 - 6.2 NeXTStation NeXTStep

SEE ALSO

The SNNS user manual under the SNNS installation directory or the hypertext manual at http://www-ra.informatik.uni-tuebingen.de/SNNS/snns.info

xgui(1) the SNNS graphical user interface

batchman(1) batch execution interface for SNNS

snns2c(1) compiler to transform a trained neural network into a C-routine

isnns(1) interactiv, on the fly neural network training for SNNS

analyze(1) analyzes result files generated by SNNS to test the classification

convert2snns(1) pattern conversion tool for Kohonen Networks

feedback-gennet(1) generator for network definition files

ff_bignet(1) complex feedforward network generator

linknets(1) connects two or more SNNS network files into one big net

mkhead(1) writes SNNS pattern file header to stdout

mkout(1) writes SNNS output pattern to stdout

mkpat(1) reads 8 bit rawfile and writes SNNS pattern file to stdout

netlearn(1) backpropagation test program

netperf(1) benchmark program

pat_sel(1) produces pattern file with selected patterns

td_bignet(1) time-delay network generator

COPYRIGHT

SNNS group of Prof. Dr. Andreas Zell, Eberhard-Karls-Universitaet Tuebingen, Koestlinstr. 6, 72074 Tuebingen, Germany, e-mail: zell@informatik.uni-tuebingen.de

BUGS

None. There are only features.

Its also a feature, that some of the man pages are not implemented yet.

AUTHOR

Guenter W. Mamier