wR1p where the symbols have their usual meanings. [8 Marks] (b) i. Show that a Newton step weight update w+1w(H1)E(w) (where H is the Hessian matrix, i.e., the (n+1)(n+1) matrix of second partial derivatives of E ), will converge to the optimal weights w in a single iteration, from any initial weight vector w0. [8 Marks] ii. Explain why this is significant, and why it might not be convenient in practice. [2 Marks]