Asked 19th Jan, 2018

Can someone provide me a reference concerning the effect of data heterogeneity on the performance of the Autoregressive model?

The major concern is how sensitive the AR model to data non homogeneity.
Kindly, I need some references.
thanks in advance

All Answers (3)

20th Jan, 2018
Ette Etuk
Rivers State University
Autoregressive (AR) modelling is based on assumption of stationarity which implies homoscedasticity and constant mean. Heteroscedasticity makes AR modelling unreliable for forecasting. To study the sensitivity of AR modelling to heteroscedasticity can best be done by simulation.
2 Recommendations
20th Jan, 2018
Adriana Santos-Caballero
University of Barcelona
21st Jan, 2018
Ayman Amin
Menoufia University
It is good if you have access to this book:
1 Recommendation

Similar questions and discussions

Related Publications

The application of computer capabilities to the technical functions of water distribution will become increasingly important as both the costs of distribution and the public pressure to reduce these costs escalate.
This article presents abstracts of papers presented at the 2009 Russian-Japanese Symposium on Stochastic Analysis of Advanced Statistical Models held from September 15 to September 17, 2009 at the Steklov Mathematical Institute of the Russian Academy of Sciences.
Got a technical question?
Get high-quality answers from experts.