I would recommend moving on one more page, to page 251 of ISLR, and using ridge regression as illustrated there instead to get your "best" ...
確定! 回上一頁