4. The adjusted-R for the linear regression model iid Y = B + BX + +k-1X k-, t + & ~ N(0,0), t = 1,2,..., T. 1t is defined as R =1- AIC* = 2k/7 / (T k) t=1 (Y-Y) / (T-1) t=1 where are the residuals from the fitted model. a. Show that selecting models to minimize AIC, SIC, or to maximize R, is the same as selecting models to minimize, respectively, T t=1 SIC* = TK/T & T T where k is number of parameters. b. Which of the three criteria in (a) results in the most parsimonious model? Which results in the least parsimonious model? 9 and R2 = 1 1-4/ T