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

Gregorova, Magda; Kalousis, Alexandros; Marchand-Maillet, Stéphane

Advanced search

Browse by...

Deposit

Highlights

More informations

Title		Learning Predictive Leading Indicators for Forecasting Time Series Systems with Unknown Clusters of Forecast Tasks
Authors		Gregorova, Magda Kalousis, Alexandros Marchand-Maillet, Stéphane
Published in		Proceedings of the Ninth Asian Conference on Machine Learning (ACML 2017). Seoul (Korea) - 15-17 November 2017 - . 2017, p. 161-176
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.
Keywords		Time series forecasting — VAR — Granger causality — Structured sparsity — Multitask learning — Leading indicators
Full text		Proceedings chapter (Published version) (483 Kb) - Free access Supplemental data (249 Kb) - Free access Other version: http://proceedings.mlr.press/v77/gregorova17a.html
Structures		Faculté des sciences / Département d'informatique Centres et instituts / Centre universitaire d'informatique
Research groups		Computer Vision and Multimedia Laboratory Viper group Geneva Artificial Intelligence Laboratory
Citation (ISO format)		GREGOROVA, Magda, KALOUSIS, Alexandros, MARCHAND-MAILLET, Stéphane. Learning Predictive Leading Indicators for Forecasting Time Series Systems with Unknown Clusters of Forecast Tasks. In: Proceedings of the Ninth Asian Conference on Machine Learning (ACML 2017). Seoul (Korea). [s.l.] : [s.n.], 2017. p. 161-176. https://archive-ouverte.unige.ch/unige:103203

65 hits

14 downloads

Contact an author

Update request

Update

Deposited on : 2018-03-26

Export document

Format	:
Citation style	:

Archive ouverte UNIGE

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