NNSYSID toolbox available

[Magnus Norgaard]

-------------------------------
ANNOUNCING: THE NNSYSID TOOLBOX
-------------------------------

                            Neural Network Based
                        System Identification Toolbox
                           for use with MATLAB(r)

                                Version 1.0

                              Magnus Norgaard
                          Institute of Automation,
                       Connect/Electronics Institute,
                     Institute of Mathematical Modeling
                       Technical University of Denmark
                               Oct. 4, 1995



The NNSYSID toolbox is a set of freeware tools for neural network based
identification of nonlinear dynamic systems. The toolbox contains a number of
m and MEX-files for training and evaluation of multilayer perceptron type
neural networks within the MATLAB environment. There are functions for
training ordinary feedforward networks as well as for identification of
nonlinear dynamic systems and for time-series analysis. The toolbox requires
at least MATLAB 4.2 with the signal processing toolbox, but it works completely
independently of the neural network and system identification toolboxes provided
by The MathWorks, Inc.



WHAT THE TOOLBOX CONTAINS:

o  Fast, robust, and easy-to-use training algorithms.
o  A number of different model structures for modelling dynamic systems.
o  Validation of trained networks.
o  Estimation of generalization ability.
o  Algorithms for determination of the optimal network architecture by pruning.
o  Demonstration programs.



HOW TO GET IT:

The toolbox can be obtained in one of the following ways:

o WWW:
     URL adress: http://kalman.iau.dtu.dk/Projects/proj/nnsysid.html
     
     zip was used for compressing the toolbox. You must have
     unzip (UNIX) or pkunzip (DOS) to unpack it. 
     
     From UNIX: unzip -a nntool.zip
     From DOS:  pkunzip nntool.zip
     

o FTP:
     ftp eivind.ei.dtu.dk
     login: anonymous
     password: (Your e-mail adress)
     cd dist/matlab/NNSYSID
     
     You will find two versions of the compressed toolbox:
     nntool.zip was packed and compressed with 'zip'
     nntool.tar.gz was packed with 'tar' and compressed with 'gzip'
     
     For the zip-version:
     get nntool.zip
     unzip -a nntool.zip (UNIX), or
     pkunzip nntool.zip (DOS)
     
     For the tar+gzip version:
     get nntool.tar.gz
     gzip -d nntool.tar.gz (UNIX only)
     tar xvf nntool.tar


After having unpacked the toolbox, read the files README and RELEASE on how to
install the tools properly. A 90-page manual (in Postscript) is included with
the toolbox.   

We do not offer any support if you run into problems! The toolbox is 
freeware - take it or leave it!!!


THE TOOLBOX FUNCTIONS GROUPED BY SUBJECT:

FUNCTIONS FOR TRAINING A NETWORK:
marq     : Levenberg-Marquardt method.
marq2    : Levenberg-Marquardt method. Works for fully connected networks only.
marqlm   : Memory-saving implementation of the Levenberg-Marquardt method.
rpe      : Recursive prediction error method.


FUNCTIONS FOR PRETREATING THE DATA:
dscale   : Scale data to zero mean and variance one.


FUNCTIONS FOR TRAINING NETWORKS TO MODEL DYNAMIC SYSTEMS:
lipschit : Determine the lag space.
nnarmax1 : Identify a Neural Network ARMAX (or ARMA) model (Linear MA filter).
nnarmax2 : Identify a Neural Network ARMAX (or ARMA) model.
nnarx    : Identify a Neural Network ARX (or AR) model.
nniol    : Identify a Neural Network model suited for I-O linearization control.
nnoe     : Identify a Neural Network Output Error model.
nnrarmx1 : Recursive counterpart to NNARMAX1.
nnrarmx2 : Recursive counterpart to NNARMAX2.
nnrarx   : Recursive counterpart to NNARX.
nnssif   : Identify a NN State Space Innovations form model.


FUNCTIONS FOR PRUNING NETWORKS:
netstruc : Extract weight matrices from matrix of parameter vectors.
nnprune  : Prune models of dynamic systems with Optimal Brain Surgeon (OBS).
obdprune : Prune feed-forward networks with Optimal Brain Damage (OBD).
obsprune : Prune feed-forward networks with Optimal Brain Surgeon (OBS).


FUNCTIONS FOR EVALUATING TRAINED NETWORKS:
fpe      : FPE estimate of the generalization error for feed-forward nets.
ifvalid  : Validation of models generated by NNSSIF.
ioleval  : Validation of models generated by NNIOL.
loo      : Leave-One-Out estimate of generalization error for feed-forward nets.
nneval   : Validation of feed-forward networks (trained by marq,rpe,bp).
nnfpe    : FPE for I/O models of dynamic systems.
nnsim    : Simulate model of dynamic system from control sequence alone.
nnvalid  : Validation of I/O models of dynamic systems.
wrescale : Rescale weights of trained network.


MISCELLANOUS FUNCTIONS:
Contents : Contents file.
drawnet  : Draws a two layer neural network.
getgrad  : Derivative of network outputs w.r.t. the weights.
pmntanh  : Fast tanh function.


DEMOS:
test1    : Demonstrates different training methods on a curve fitting example.
test2    : Demonstrates the NNARX function.
test3    : Demonstrates the NNARMAX2 function.
test4    : Demonstrates the NNSSIF function.
test5    : Demonstrates the NNOE function.
test6    : Demonstrates the effect of regularization by weight decay.
test7    : Demonstrates pruning by OBS on the sunspot benchmark problem.


Enjoy
- MN

  
+-----------------------------------+------------------------------+
| Magnus Norgaard                   | e-mail : pmn at iau.dtu.dk      |
| Institute of Automation           | Phone  : +45 42 25 35 65     |
| Technical University of Denmark   | Fax    : +45 45 88 12 95     |
| Building 326                      | http://kalman.iau.dtu.dk/    |
| DK-2800 Lyngby                    |    Staff/Magnus_Norgaard.html|
| Denmark                           |                              |
|___________________________________|______________________________|