Consider the model, defined by the following assumptions: Al (Linearity): Y = XB + U, A2 (Exogeneity): E(U X) = 0 or A2* : E(U;Xi) = 0 and E(Ui) = 0, for i = 1, ..., n, A3 (Homoskedasticity): Var(U|X) = 02In, A4 (No Multicollinearity): rank(X) = k, where X is an n x k matrix A5: UX ~ N(0, 02 In). a) [4 points Derive the least square estimator of B, i.e., BOLS. b) [4 points] Show that OLS estimator is unbiased. c) [4 points] Propose an unbiased estimator of o2. Show that under usual assumptions (Al) - (A4), this proposed estimator is unbiased. d) [3 points] Describe the distribution of U'U/o2