(ii) Consider the linear model yi = x; B+6, i = 1,2. . ... n, (3) where & is an i id. sequence of random variables with zero mean and variance Var(e) =o'q, for some variance o' and q > 0. Discounted least squares considers the maximum likelihood estimator 3 of B, which minimises the discounted sum of squares S.(B) = (-x8), for some discount factor & that satisfies 0