Comparative performance of autoregressive order determination criteria for subset modelling

  • EH Etuk

Abstract

The efficiencies of eight autoregressive model order determination criteria:
AIC, BIC, SIC, S, F, FPE4, CAT2, CAT3, for the selection of subset models are
compared using artificial and real series. Our observation is that BIC, S and FPE4 performed well in a wide range of models.
Section
Articles

Journal Identifiers


eISSN: 1116-4336