3. (10 pts) Suppose we are given a neural network with an architecture as in problem 1 and we want to train the network through a set of 300 example inputs and their corresponding outputs to calculate the unknown vector-valued function y=[y1,y2] for all possible inputs x=(x1,x2,,x20). Can we apply the gradient descent algorithm to find the optimal values for all the weights and biases of the neurons in this neural network using the training set? Why or why not? What about applying the stochastic gradient descent algorithm instead