Abstract: Model selection methods are developed using the approach of hypothesis testing. The technique of cross-validation has been used to make the procedures fully automated. Simulation results are presented to show that the proposed methods work better than Akaike's FPE, Mallows' Cp or Schwartz's BIC.
Key words and phrases: Hypothesis testing, model selection, Akaike's FPE, Mallows' Cp, Schwartz's BIC.