2. A scientist has measured expression levels for 5,[H]l] gene: on 180 patients) He is interested in building a. negremion model based upon the genes to predict a continuous clinical response and decides to use the LASSO. He found that it is neceesarjr to standardize the gene expression valuai He implements the following procedure to build his model: 1. Standardize all gene features such that each feature has zero sample mean and unit sample standard deviation, computed with 130 data. Note that the mponse is not standardized. 2. l-iirnplog.r 5fold crowvalidation on the standardized 5'000 genes to select the tuning parameter [A] in the LASSC' and report the prediction error. Questions: {a} Is this procedure correct and raults in a sound estimate of the prediction error? Why? {b} Howr would you change in this procedure? Please write a pseudo-code to elaborate your new procedure. Suppose that the range of A is given by grid=13nseqf5r| 3,engh=50j