Compute the gradient of the weights W = (w, W2, W3, w4) for the following neural network and loss objective (L). You should give a final formula of the gradient as a function of x = (x1, x2) and y. Here o is the sigmoid function. Compute: h(W, x) = w0 (w3x1 + W4x2) + W20(W3x1 -W4x2) L(W, x, y) = (y - h(W, x)) Vw L(W, x, 7, x, y)