Use Different Criteria for Model Selection
Select a time series model for data based on different selection criteria, such as Akaike information criterion (AIC), finite sample corrected AIC, Bayesian information criterion (BIC), or Schwarz–Bayes information criterion (SBC).
Display the table of the ranked candidate models for each criterion.
Select the top three entries from each table and refine them using the maximum likelihood estimation. Models estimated with different methods may have different rankings.
Define ranking function for each criterion.
Select the best maximum likelihood estimates according to different criteria.