machine learning questions help neural network and explain please

thank you

Given the neural network below: Assume that the biases are fixed at zero and not allowed to change. Also, assume that the activation functions are the identity, and therefore the neurons are linear. The weight values in the figure indicate the initial values. It is intended to train this network for one iteration, with a dataset consisting of only one feature vector: (x1=1;2=1) and the corresponding label 01=0. Question 15 The number of hidden layers is Perform forward propagation for the network. Indicate the numerical values of all hidden neurons and of the output neuron. h1= h22= 0_1= Question 17 The squared error of the network with the given dataset is Question 18 Perform back propagation for the network. Indicate the gradient of the weights with respect to the loss (squared error). There are 6 numeric values to compute ( 6 weights). Derivative of L with respect to w11(1) : Derivative of L with respect to w_12 (1) = Derivative of L with respect to w221(1)= Derivative of L with respect to w_22 (1) = Derivative of L with respect to w11(2)= Derivative of L with respect to w12 (2) = Perform one step of gradient descent for the network. Assume that the step size is 0.01 and there is no momentum. Your goal is to report the new values of the weights. w11(1)=w12(1)=w21(1)=w22(1)=w11(2)=w12(2)=