Optimal Estimation of Multivariate ARMA Models

Martha White, Junfeng Wen, Michael Bowling, Dale Schuurmans

AAAI 2015 pp. 3080-3086

doi:10.1609/AAAI.V29I1.9614 /aaai/2015/white2015aaai-optimal/

Abstract

Autoregressive moving average (ARMA) models are a fundamental tool in timeseries analysis that offer intuitive modeling capability and efficient predictors. Unfortunately, the lack of globally optimal parameter estimation strategies for these models remains a problem:application studies often adopt the simpler autoregressive model that can be easily estimated by maximizing (a posteriori) likelihood. We develop a (regularized, imputed) maximum likelihood criterion that admits efficient global estimation via structured matrix norm optimization methods. An empirical evaluation demonstrates the benefits of globally optimal parameter estimation over local and moment matching approaches.

PDF AAAI Semantic Scholar

Cite

Text

White et al. "Optimal Estimation of Multivariate ARMA Models." AAAI Conference on Artificial Intelligence, 2015. doi:10.1609/AAAI.V29I1.9614

Markdown

[White et al. "Optimal Estimation of Multivariate ARMA Models." AAAI Conference on Artificial Intelligence, 2015.](https://mlanthology.org/aaai/2015/white2015aaai-optimal/) doi:10.1609/AAAI.V29I1.9614

BibTeX

@inproceedings{white2015aaai-optimal,
  title     = {{Optimal Estimation of Multivariate ARMA Models}},
  author    = {White, Martha and Wen, Junfeng and Bowling, Michael and Schuurmans, Dale},
  booktitle = {AAAI Conference on Artificial Intelligence},
  year      = {2015},
  pages     = {3080-3086},
  doi       = {10.1609/AAAI.V29I1.9614},
  url       = {https://mlanthology.org/aaai/2015/white2015aaai-optimal/}
}