6. Time Series Model Selection

  1. Søren Bisgaard and
  2. Murat Kulahci

Published Online: 31 MAY 2011

DOI: 10.1002/9781118056943.ch6

Time Series Analysis and Forecasting by Example

Time Series Analysis and Forecasting by Example

How to Cite

Bisgaard, S. and Kulahci, M. (2011) Time Series Model Selection, in Time Series Analysis and Forecasting by Example, John Wiley & Sons, Inc., Hoboken, NJ, USA. doi: 10.1002/9781118056943.ch6

Author Information

  1. Technical University of Denmark

Publication History

  1. Published Online: 31 MAY 2011
  2. Published Print: 5 JUL 2011

Book Series:

  1. Wiley Series in Probability and Statistics

Book Series Editors:

  1. Walter A. Shewhart and
  2. Samuel S. Wilks

ISBN Information

Print ISBN: 9780470540640

Online ISBN: 9781118056943

SEARCH

Keywords:

  • Akaike's bias-corrected information criterion (AICC);
  • Akaike's information criterion (AIC);
  • Bayesian information criterion (BIC);
  • impulse response functions;
  • maximum likelihood estimation;
  • time series model

Summary

One way to think of a time series model is that it is an approximation to some unknown dynamic system. Akaike’s information criterion (AIC) is used for model selection, especially in the time series context. The author uses an example to discuss the problem of model selection and the use of model selection criteria. The most popular criteria are Akaike’s information criterion (AIC), Akaike's bias-corrected information criterion (AICC) suggested by Hurvich and Tsai, and the Bayesian information criterion (BIC) introduced by Schwarz. One approach to time series modeling is to fit a number of potential autoregressive moving average (ARMA) models to the data using the maximum likelihood estimation, choose a criterion, and select the model that has the best value according to this criterion. Impulse response functions are usually computed for stationary models.

Controlled Vocabulary Terms

Bayesian information criterion; maximum likelihood estimation; time series analysis