Q1 (50%) : Suppose we observe a training set thai (1, ail,4a, . . . , xip)) for i-1, , , , , n subjects. We want to fit a linear regression model f(x)-. Atyv 1 2jÃi to the training data to minimize the residual sum of squares RSS(β) -Σ η 1 (Vi-zP)2 where β = (As,å±, , Ãp)T (d)(10%) Is it possible to use cross validation to choose the optimal tuning parameter λ such that the resulting model f(x) Ãridge + Σ-1x,Bridge achieves both the minimum bias and minimum variance? If yes, please explain how to conduct the cross validation. If not, please explain why. (e)(10%) Please write out the optimization problem for SVM regression. What's the difference between SVM regression and ridge regression?