Learning Predictive Leading Indicators for Forecasting Time Series Systems with Unknown Clusters of Forecast Tasks

Magda Gregorová, Alexandros Kalousis, Stéphane Marchand-Maillet
Proceedings of the Ninth Asian Conference on Machine Learning, PMLR 77:161-176, 2017.

Abstract

We present a new method for forecasting systems of multiple interrelated time series. The method learns the forecast models together with discovering leading indicators from within the system that serve as good predictors improving the forecast accuracy and a cluster structure of the predictive tasks around these. The method is based on the classical linear vector autoregressive model (VAR) and links the discovery of the leading indicators to inferring sparse graphs of Granger causality. We formulate a new constrained optimisation problem to promote the desired sparse structures across the models and the sharing of information amongst the learning tasks in a multi-task manner. We propose an algorithm for solving the problem and document on a battery of synthetic and real-data experiments the advantages of our new method over baseline VAR models as well as the state-of-the-art sparse VAR learning methods.

Cite this Paper

BibTeX
@InProceedings{pmlr-v77-gregorova17a, title = {Learning Predictive Leading Indicators for Forecasting Time Series Systems with Unknown Clusters of Forecast Tasks}, author = {Gregorová, Magda and Kalousis, Alexandros and Marchand-Maillet, Stéphane}, booktitle = {Proceedings of the Ninth Asian Conference on Machine Learning}, pages = {161--176}, year = {2017}, editor = {Zhang, Min-Ling and Noh, Yung-Kyun}, volume = {77}, series = {Proceedings of Machine Learning Research}, address = {Yonsei University, Seoul, Republic of Korea}, month = {15--17 Nov}, publisher = {PMLR}, pdf = {http://proceedings.mlr.press/v77/gregorova17a/gregorova17a.pdf}, url = {https://proceedings.mlr.press/v77/gregorova17a.html}, abstract = {We present a new method for forecasting systems of multiple interrelated time series. The method learns the forecast models together with discovering leading indicators from within the system that serve as good predictors improving the forecast accuracy and a cluster structure of the predictive tasks around these. The method is based on the classical linear vector autoregressive model (VAR) and links the discovery of the leading indicators to inferring sparse graphs of Granger causality. We formulate a new constrained optimisation problem to promote the desired sparse structures across the models and the sharing of information amongst the learning tasks in a multi-task manner. We propose an algorithm for solving the problem and document on a battery of synthetic and real-data experiments the advantages of our new method over baseline VAR models as well as the state-of-the-art sparse VAR learning methods.} }
Endnote
%0 Conference Paper %T Learning Predictive Leading Indicators for Forecasting Time Series Systems with Unknown Clusters of Forecast Tasks %A Magda Gregorová %A Alexandros Kalousis %A Stéphane Marchand-Maillet %B Proceedings of the Ninth Asian Conference on Machine Learning %C Proceedings of Machine Learning Research %D 2017 %E Min-Ling Zhang %E Yung-Kyun Noh %F pmlr-v77-gregorova17a %I PMLR %P 161--176 %U https://proceedings.mlr.press/v77/gregorova17a.html %V 77 %X We present a new method for forecasting systems of multiple interrelated time series. The method learns the forecast models together with discovering leading indicators from within the system that serve as good predictors improving the forecast accuracy and a cluster structure of the predictive tasks around these. The method is based on the classical linear vector autoregressive model (VAR) and links the discovery of the leading indicators to inferring sparse graphs of Granger causality. We formulate a new constrained optimisation problem to promote the desired sparse structures across the models and the sharing of information amongst the learning tasks in a multi-task manner. We propose an algorithm for solving the problem and document on a battery of synthetic and real-data experiments the advantages of our new method over baseline VAR models as well as the state-of-the-art sparse VAR learning methods.
APA
Gregorová, M., Kalousis, A. & Marchand-Maillet, S.. (2017). Learning Predictive Leading Indicators for Forecasting Time Series Systems with Unknown Clusters of Forecast Tasks. Proceedings of the Ninth Asian Conference on Machine Learning, in Proceedings of Machine Learning Research 77:161-176 Available from https://proceedings.mlr.press/v77/gregorova17a.html.

Related Material