This question tries to understand the connection between test loss and classification accuracy or these models. Part(a) : Show that if the average of N non-negative numbers is , and k is a positive integer, then