Computer science exercise:

Neural Network with Computer Vision

Starting from the provided code in implement the learning of a Neural Network with 3 layers
of the following size:
2 input units
4 hidden units
1 output unit
Note: the 2 and 4 exclude the bias unit but your network should be implemented with a bias unit on
both the input layer and the hidden layer.
Implement functions for sigmoid and sigmoidprime.
Code in is provided that reads in 200 training examples into desMat and ytrain.



Powered by WordPress