LMS regression can be solved analytically. Given a data set D = {(x1, y1), (x2, y2), ..., (Xn, Un)}, define a matrix X E Rnxm and vector y E R" as follows: X1 y1 y2 X = y : . . . yn where each sample x; has m features. Show that the optimization problem for LMS regression is equivalent to minw (Xw -y) (Xw -y), and that the solution w* is w* = (XTX)-iXTy