2. Consider the Normal regression model which is partly linear and partly non- linear:
yi D þ1 C þ2 xi 2 CÐ Ð Ð C þk xik C f .xi;kC1; ::: ; xi;kCp / C "i
(a) Using a noninformative prior, derive the marginal posterior p jy/.
(b) Discuss how you might derive a posterior simulator if you wished to carry out Bayesian inference solely on .
(c) Discuss how you might derive a posterior simulator if you wished to carry out Bayesian inference on and þ1;:: : ; þk .