In the context of multiple linear regression, let Y be the nxl vector contain the responses, let X be the nx{p+1) array whose rst column consists entirely of ones and whose other columns are the p explanatory variable values, and let b be the (p+1)x1 column containing the resulting parameter estimates. We know that as long as the matrix inversion is possible, the least squares solution is b = [rm11:7? 1. When is possible to find the inverse of the matrix A\"! ? 2. Show that EH1] = 3. Show that Cov[h] = 02(XTX}_1 [Hint: IfA nxn is a matrix and Y is a nxl vector with mean pm] and Covariance 2, then EMF] = A u and supply) = A CwmAT ]