Online Learning for Time Series Prediction
Abstract
We address the problem of predicting a time series using the ARMA (autoregressive moving average) model, under minimal assumptions on the noise terms. Using regret minimization techniques, we develop eective online learning algorithms for the prediction problem, without assuming that the noise terms are Gaussian, identically distributed or even independent. Furthermore, we show that our algorithm’s performances asymptotically approaches the performance of the best ARMA model in hindsight.
Cite
Text
Anava et al. "Online Learning for Time Series Prediction." Annual Conference on Computational Learning Theory, 2013.Markdown
[Anava et al. "Online Learning for Time Series Prediction." Annual Conference on Computational Learning Theory, 2013.](https://mlanthology.org/colt/2013/anava2013colt-online/)BibTeX
@inproceedings{anava2013colt-online,
title = {{Online Learning for Time Series Prediction}},
author = {Anava, Oren and Hazan, Elad and Mannor, Shie and Shamir, Ohad},
booktitle = {Annual Conference on Computational Learning Theory},
year = {2013},
pages = {172-184},
url = {https://mlanthology.org/colt/2013/anava2013colt-online/}
}