Ensembles for Time Series Forecasting

Mariana Oliveira, Luis Torgo
Proceedings of the Sixth Asian Conference on Machine Learning, PMLR 39:360-370, 2015.

Abstract

This paper describes a new type of ensembles that aims at improving the predictive performance of these approaches in time series forecasting. Ensembles are recognised as one of the most successful approaches to prediction tasks. Previous theoretical studies of ensembles have shown that one of the key reasons for this performance is diversity among ensemble members. Several methods exist to generate diversity. The key idea of the work we are presenting here is to propose a new form of diversity generation that explores some specific properties of time series prediction tasks. Our hypothesis is that the resulting ensemble members will be better at addressing different dynamic regimes of time series data. Our large set of experiments confirms that the methods we have explored for generating diversity are able to improve the performance of the equivalent ensembles with standard diversity generation procedures.

Cite this Paper


BibTeX
@InProceedings{pmlr-v39-oliveira14, title = {Ensembles for Time Series Forecasting}, author = {Oliveira, Mariana and Torgo, Luis}, booktitle = {Proceedings of the Sixth Asian Conference on Machine Learning}, pages = {360--370}, year = {2015}, editor = {Phung, Dinh and Li, Hang}, volume = {39}, series = {Proceedings of Machine Learning Research}, address = {Nha Trang City, Vietnam}, month = {26--28 Nov}, publisher = {PMLR}, pdf = {http://proceedings.mlr.press/v39/oliveira14.pdf}, url = {https://proceedings.mlr.press/v39/oliveira14.html}, abstract = {This paper describes a new type of ensembles that aims at improving the predictive performance of these approaches in time series forecasting. Ensembles are recognised as one of the most successful approaches to prediction tasks. Previous theoretical studies of ensembles have shown that one of the key reasons for this performance is diversity among ensemble members. Several methods exist to generate diversity. The key idea of the work we are presenting here is to propose a new form of diversity generation that explores some specific properties of time series prediction tasks. Our hypothesis is that the resulting ensemble members will be better at addressing different dynamic regimes of time series data. Our large set of experiments confirms that the methods we have explored for generating diversity are able to improve the performance of the equivalent ensembles with standard diversity generation procedures.} }
Endnote
%0 Conference Paper %T Ensembles for Time Series Forecasting %A Mariana Oliveira %A Luis Torgo %B Proceedings of the Sixth Asian Conference on Machine Learning %C Proceedings of Machine Learning Research %D 2015 %E Dinh Phung %E Hang Li %F pmlr-v39-oliveira14 %I PMLR %P 360--370 %U https://proceedings.mlr.press/v39/oliveira14.html %V 39 %X This paper describes a new type of ensembles that aims at improving the predictive performance of these approaches in time series forecasting. Ensembles are recognised as one of the most successful approaches to prediction tasks. Previous theoretical studies of ensembles have shown that one of the key reasons for this performance is diversity among ensemble members. Several methods exist to generate diversity. The key idea of the work we are presenting here is to propose a new form of diversity generation that explores some specific properties of time series prediction tasks. Our hypothesis is that the resulting ensemble members will be better at addressing different dynamic regimes of time series data. Our large set of experiments confirms that the methods we have explored for generating diversity are able to improve the performance of the equivalent ensembles with standard diversity generation procedures.
RIS
TY - CPAPER TI - Ensembles for Time Series Forecasting AU - Mariana Oliveira AU - Luis Torgo BT - Proceedings of the Sixth Asian Conference on Machine Learning DA - 2015/02/16 ED - Dinh Phung ED - Hang Li ID - pmlr-v39-oliveira14 PB - PMLR DP - Proceedings of Machine Learning Research VL - 39 SP - 360 EP - 370 L1 - http://proceedings.mlr.press/v39/oliveira14.pdf UR - https://proceedings.mlr.press/v39/oliveira14.html AB - This paper describes a new type of ensembles that aims at improving the predictive performance of these approaches in time series forecasting. Ensembles are recognised as one of the most successful approaches to prediction tasks. Previous theoretical studies of ensembles have shown that one of the key reasons for this performance is diversity among ensemble members. Several methods exist to generate diversity. The key idea of the work we are presenting here is to propose a new form of diversity generation that explores some specific properties of time series prediction tasks. Our hypothesis is that the resulting ensemble members will be better at addressing different dynamic regimes of time series data. Our large set of experiments confirms that the methods we have explored for generating diversity are able to improve the performance of the equivalent ensembles with standard diversity generation procedures. ER -
APA
Oliveira, M. & Torgo, L.. (2015). Ensembles for Time Series Forecasting. Proceedings of the Sixth Asian Conference on Machine Learning, in Proceedings of Machine Learning Research 39:360-370 Available from https://proceedings.mlr.press/v39/oliveira14.html.

Related Material