C++ Neural Networks and Fuzzy Logic: Preface

C++ Neural Networks and Fuzzy Logic

by Valluru B. Rao

MTBooks, IDG Books Worldwide, Inc.

ISBN: 1558515526   Pub Date: 06/01/95

Previous Table of Contents Next

A Source File for C++ Program for an ART1 Model Network

The implementations of the functions declared in the header file are contained in the source file for the C++

program for an ART1 model network. It also has the main function, which contains specifications of the

number of neurons in the two layers of the network, the values of the vigilance and other parameters, and the

input vectors. Note that if there are n neurons in a layer, they are numbered serially from 0 to n–1, and not

from 1 to n in the C++ program. The source file is called art1net.cpp. It is set up with six neurons in the F

1

layer and seven neurons in the F

2

 layer. The main function also contains the parameters needed in the

algorithm.

To initialize the bottom−up weights, we set each weight to be –0.1 + L/(m – 1 + L) so that it is greater than 0

and less than L/(m – 1 + L), as suggested before. Similarly, the top−down weights are initialized by setting

each of them to 0.2 + (B – 1)/D so it would be greater than (B – 1)/D. Initial activations of the F

1

 layer

neurons are each set to –B/(1 + C), as suggested earlier.

A restrmax function is defined to compute the maximum in an array when one of the array elements is not

desired to be a candidate for the maximum. This facilitates the removal of the current winner from

competition when reset is needed. Reset is needed when the degree of match is of a smaller magnitude than

the vigilance parameter.

The function iterate is a member function of the network class and does the processing for the network. The