OVERVIEW

A simple implementation of a Back-Propagation Artificial Neural Network in C.

COMPILE

make

OR

gcc -c *.c

gcc -o backprop backprop.o ez_alloc.o datafile.o rand.o -lm

USAGE

backprop <data file> <wts file> <0=train/1=recall> [<0=logistic/1=hyperbolic tangent>] [<number of outputs>]

EXAMPLES

backprop xor.dat xor.wts 0 0 1

backprop xor_recall.dat xor.wts 1 0 1

backprop iris.dat iris.wts 0 0 3

backprop iris_recall.dat iris.wts 1 0 3

REFERENCES

Theory for implementation provided by "Neural Networks : A Comprehensive Foundation" by Simon Haykin. (Prentice Hall, Second Edition, 1999)