Returning to the general form of a Poisson regression loss function () = = 1[ ln ()] and applying a typical GLM structure for Poisson regression with one predictor variable : = exp (0 + 1) a. Restate the loss function as a function of = (0,1) rather than . b. Take the partial derivative of the new function () (i.e. (0,1)) with respect to 0. c. Show that, by setting the expression from part b equal to 0, an optimal choice of must force the mean of all modeled values 1... to be equal to the mean of all observed values 1...