In Example 3.6 use the WinBUGS jump RJMCMC interface to obtain the highest posterior probability model for the Chevrolet asking price data, assuming a prior \(P_{\text {ret }} \sim\) \(\operatorname{Bin}(P, 0.5)\) on the number of retained predictors (with \(P=5\) ). This can be implemented by adapting the RJMCMC code from Example 3.3, including statements for the conjugate normal prior assumed in Example 3.6:

How far does this model remedy the predictive deficiencies identified in Example 3.6.

Data from Example 3.6

Transcribed Image Text:

tau.beta.prec