Bayesian model selection for linear regression

In this note we introduce linear regression with basis functions in order to apply Bayesian model selection. The goal is to incorporate Occam’s razor as provided by Bayes analysis in order to automatically pick the model optimally able to explain the data without overfitting.