(a) Let X1, . .., Xn be a sample of real valued data, and let Xn = _ _ _, X, denote the sample mean. Show that n Xn = arg min aER [(x; - a)? j=1 (b) Now consider a sample X1, ..., Xn of d-dimensional feature vectors, d > 2, and let Xn be the sample mean (also a d-dimensional vector). Show that n Xn = arg min >IX, - all2. aERd j=1