A neural network for a two-class classification problem is given below. The units at the hidden layer have ReLU activations.

The weights determined through training are:

w11=-1.5; w12=-1.0, w21=1.2; w22=1.0; w31=-0.6; w32=1.0 h11=-0.5; h12=-1.0; h21=1.0, h22=1.5; h31= 0.4; h32=0.3

Suppose x1=4, and x2=5.

(a) Calculate the net inputs a1 and a2.

(b) Calculate the ReLU outputs.

(c) Calculate y1 and y2.

(d) Calculate z1 and z2.

(e) If cross entropy used as the loss function, and the target value is [1,0], calculate the loss