19.8 Consider the regression model Yi = b0 + b1Xi + ui, where u1 = u 

1 and ui = 0.5ui - 1 + u 

i for i = 2, 3,

c, n. Suppose that u 

i are i.i.d. with mean 0 and variance 1 and are distributed independently of Xj for all i and j.

a. Derive an expression for E(U U) = .

b. Explain how to estimate the model by GLS without explicitly inverting the matrix . (Hint: Transform the model so that the regression errors are u  1, u  2,

c, u  n.)