For the Bayesian ordinary normal linear model, using a flat improper prior for ???? but a proper inverse-gamma distribution (10.5) for ????2, find the posterior Bayes estimate of ????2 and express it as a weighted average of s2 and the prior mean of ????2.