The following neural network consists of three layer (layer 1: input layer, layer 2: hidden layer, layer3: output layer). Where x = [2 3 1]' is the input vector, the weight vectors is: -3.5 (1) = = [ { -2] (2) = 0.5 -1.2 10.3 0.6 Assume each unit has a bias 0 =0 (zero) 1 The activation units is calculated by sigmoid function g(z) (where in your 1 +exp(-2) calculator use ex i.e. consider base 2 but not use exp) X1 X2 a. Calculate af of the second hidden layer b. Calculate as of the output layer