Aggregating strategies for long-term forecasting

Alexander Korotin, Vladimir V’yugin, Evgeny Burnaev
Proceedings of the Seventh Workshop on Conformal and Probabilistic Prediction and Applications, PMLR 91:63-82, 2018.

Abstract

The article is devoted to investigating an application of aggregating algorithms to the problem of the long-term forecasting. We examine the classic aggregating algorithms based on the exponential reweighing. For the general Vovk’s aggregating algorithm we provide its probabilistic interpretation and its generalization for the long-term forecasting. For the special basic case of Vovk’s algorithm we provide two its modifications for the long-term forecasting. The first one is theoretically close to an optimal algorithm and is based on replication of independent copies. It provides the time-independent regret bound with respect to the best expert in the pool. The second one is not optimal but is more practical (explicitly models dependencies in observations) and has $O(\sqrtT)$ regret bound, where $T$ is the length of the game.

Cite this Paper


BibTeX
@InProceedings{pmlr-v91-korotin18a, title = {Aggregating strategies for long-term forecasting}, author = {Korotin, Alexander and V’yugin, Vladimir and Burnaev, Evgeny}, booktitle = {Proceedings of the Seventh Workshop on Conformal and Probabilistic Prediction and Applications}, pages = {63--82}, year = {2018}, editor = {Gammerman, Alex and Vovk, Vladimir and Luo, Zhiyuan and Smirnov, Evgueni and Peeters, Ralf}, volume = {91}, series = {Proceedings of Machine Learning Research}, month = {11--13 Jun}, publisher = {PMLR}, pdf = {http://proceedings.mlr.press/v91/korotin18a/korotin18a.pdf}, url = {https://proceedings.mlr.press/v91/korotin18a.html}, abstract = {The article is devoted to investigating an application of aggregating algorithms to the problem of the long-term forecasting. We examine the classic aggregating algorithms based on the exponential reweighing. For the general Vovk’s aggregating algorithm we provide its probabilistic interpretation and its generalization for the long-term forecasting. For the special basic case of Vovk’s algorithm we provide two its modifications for the long-term forecasting. The first one is theoretically close to an optimal algorithm and is based on replication of independent copies. It provides the time-independent regret bound with respect to the best expert in the pool. The second one is not optimal but is more practical (explicitly models dependencies in observations) and has $O(\sqrtT)$ regret bound, where $T$ is the length of the game.} }
Endnote
%0 Conference Paper %T Aggregating strategies for long-term forecasting %A Alexander Korotin %A Vladimir V’yugin %A Evgeny Burnaev %B Proceedings of the Seventh Workshop on Conformal and Probabilistic Prediction and Applications %C Proceedings of Machine Learning Research %D 2018 %E Alex Gammerman %E Vladimir Vovk %E Zhiyuan Luo %E Evgueni Smirnov %E Ralf Peeters %F pmlr-v91-korotin18a %I PMLR %P 63--82 %U https://proceedings.mlr.press/v91/korotin18a.html %V 91 %X The article is devoted to investigating an application of aggregating algorithms to the problem of the long-term forecasting. We examine the classic aggregating algorithms based on the exponential reweighing. For the general Vovk’s aggregating algorithm we provide its probabilistic interpretation and its generalization for the long-term forecasting. For the special basic case of Vovk’s algorithm we provide two its modifications for the long-term forecasting. The first one is theoretically close to an optimal algorithm and is based on replication of independent copies. It provides the time-independent regret bound with respect to the best expert in the pool. The second one is not optimal but is more practical (explicitly models dependencies in observations) and has $O(\sqrtT)$ regret bound, where $T$ is the length of the game.
APA
Korotin, A., V’yugin, V. & Burnaev, E.. (2018). Aggregating strategies for long-term forecasting. Proceedings of the Seventh Workshop on Conformal and Probabilistic Prediction and Applications, in Proceedings of Machine Learning Research 91:63-82 Available from https://proceedings.mlr.press/v91/korotin18a.html.

Related Material