Learning Predictive Leading Indicators for Forecasting Time Series Systems with Unknown Clusters of Forecast Tasks
This work addresses forecasting in systems with unknown clusters of tasks, offering incremental improvements for time series analysis.
The authors tackled the problem of forecasting multiple interrelated time series by developing a method that learns forecast models while discovering leading indicators and cluster structures, based on a linear vector autoregressive model and sparse Granger causality graphs. They reported advantages over baseline and state-of-the-art methods in synthetic and real-data experiments.
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.