ahmedshoaib / backpropagation Goto Github PK

This program implements the Back propagation algorithm on Multi-layer perceptron (1 hidden layer, n hidden neurons) in incremental mode using sigmoidal activation function.

Description of input files:

Line 1 : n                                 //no of attributes
Line 2 : x1 x2 x3 .... xn y11 y12 ....y1n               //Training sample 1  y is binary classification where  '1 0 0 ' indicates class 1
Line 3 : x1 x2 x3 .... xn y21 y22 .... y2n
.
.
.
Line m : x1 x2 x3 .... xn ym1 ym2 ......ymn         //Training sample m

test sample -- enter test attributes :
x1
x2
.
.
xn