(5 points) Consider a simple two-layer network in the lecture slides. Given m train- ing data (x, y'), i = 1,..., m, the cost function used to training the neural networks m (( w, a, B) = > (y' - o( w ] z ?) ) 2 i=1 where o(x) = 1/(1 +e-") is the sigmoid function, 2' is a two-dimensional vector such that z = o(a x? ), and 23 = o(31x ). Show the that the gradient is given by Ole( w, a, B) m Ow _2(y' - o(u'))o(u') (1 - o(u'))2, i=1 where u' = w/z'. Also find the gradient of ((w, a, B) with respect to a and B and write down their expression