ABSTRACT
This chapter discusses the model selection and model checking, which are special cases of model assessment. It discusses hypothesis testing, which is a special case of model selection. Bayesian hypothesis testing involves a formal decision-making process to choose between two hypotheses. With a fixed number of models, and when proper priors are used, the most straightforward Bayesian approach simply selects the model associated with the largest posterior probability. The chapter discusses the methods of model selection that are based on how well a model might predict future observations. Historically, one of the most important model selection criteria has been the Bayesian information criterion, also called the “Schwarz information criterion,” proposed by Schwarz.
