It started in my junior high and now i want it to atleast behave like a neural net engine or do some stuff
any contributions ?? feel free to add them
INPUT LAYER | HIDDEN LAYER | OUTPUT LAYER |
---|---|---|
(N1,W1) | (N1,W1)..(N1,W1) | |
(N2,W2) | (N2,W2)..(N2,W2) | OUTPUT |
(N3,W3) | ............. | |
...... | ............ | |
...... | (N,W)..(N,W) | |
(N,W)__ | _________________________ | ___________________________ |
OUTPUT = SIGMOID(N1W1 + N2W2 + .. + Nn*Wn) | ||
Calculates error in the Weights from gradient descendant method | ||
if desired output && desired error has been reached it stops updating the Weights and conclude that the Network has been trained much |