Expert Q&A .) Given a set of N triplets {xX,y), 1 sis N, the goal is to design a model to predict y based on the features X and X. For this, we use the following neural network. In this model, the weights W., W.,..., W. and bi - fws W2 1:6-fo po D) w b3 single neuron W3 bz W6 single neuron W4 +-FO single neuron Hidden Layer the biases bl. bz, by should be learned. The (nonlinear) activation function f() is defined as (positive integer q isa network model parameter which is fixed in advance): f(z) = (2z + Iz!) The mean squared error of the network is measured by - (W.W3.....W., bu, bz, b) =- (ym) - ((W..W.. ....We, bz, bz, bx))* a) Write down the gradients of the mean squared error E(WW2, ... Wc.b. 63, b) with respect to all the parameters. Show an outline of your derivation you do not need to compute the exact derivatives, but sufficiently describe the outline). b) Describe a gradient descent algorithm to estimate the parameters. c) For q = 1, can you derive an equivalent but simpler neural network (Le., a network without a hidden layer)? Prove your answer. d) Forq = 1, is the model equivalent to a linear regression model? Explain your