Question 5* [-l8=12] We wish to predict the response 1"\" at a mmate vector 11 in a general linear model Y 2 X3 + e. The predicted value at u is YA" = ur where 5' denotes the least squares estimate of ,3 based upon the data (Y, K} . The prediction error is dened to he E = Y' "i". 5.1 Show, taking care to justify a.\" the steps in your derivation, that serfs} = {l + ur[X'X]_lu]-og (1} where Hg is the eoninlon variance of the components of the error vector 2. 5.2 Show that formula [2.13] on page 243 is a special case of [1}. In other words, show that if you substitute u=[1x*]'m1d3r=[ 1 I: 1] into {1], then you get