Layer
- input layer
- output layer
- hidden layer
%7D_i)
is "activation" of unit i in layer j
is matrix of weights controlling function mapping from layer j to layer j+1
%7Dx_0%2B%5CTheta_%7B11%7D%5E%7B(1)%7Dx_1%2B%5CTheta_%7B12%7D%5E%7B(1)%7Dx_2%2B%5CTheta_%7B13%7D%5E%7B(1)%7Dx_3))
%7Dx_0%2B%5CTheta_%7B21%7D%5E%7B(1)%7Dx_1%2B%5CTheta_%7B22%7D%5E%7B(1)%7Dx_2%2B%5CTheta_%7B23%7D%5E%7B(1)%7Dx_3))
%7Dx_0%2B%5CTheta_%7B31%7D%5E%7B(1)%7Dx_1%2B%5CTheta_%7B32%7D%5E%7B(1)%7Dx_2%2B%5CTheta_%7B33%7D%5E%7B(1)%7Dx_3))
%3Dg(%5CTheta_%7B10%7D%5E%7B(1)%7Da_0%2B%5CTheta_%7B11%7D%5E%7B(1)%7Da_1%2B%5CTheta_%7B12%7D%5E%7B(1)%7Da_2%2B%5CTheta_%7B13%7D%5E%7B(1)%7Da_3))
if network has 
units in layer j,
units in layer j+1,then %7D)
will be of dimension )
XNOR implement use Neural networks
multi-class classfication use Neural networks
