i have a few quesitons about neural networks.
i have a sample structure of one if you need it, just ask me for it,
if it's easier to just show an example thanks.

a)given the neural network structure with inputs, weights, and
activation functions how do i figure out the expression for the output

b)what restrictions would the absence of biases place upon the class
of functions a3 = y(x1, x2) where a3 is the output to the network
which can be realized without error

c)how do i calculate output error

d)how do i figure out sensitivity matrices for backpropagation