STEMO: Early Spatio-temporal Forecasting with Multi-Objective Reinforcement Learning

TL;DR

This paper introduces STEMO, a multi-objective reinforcement learning model that balances accuracy and timeliness for early spatio-temporal forecasting, crucial for applications like wildfire and traffic prediction.

Contribution

STEMO is the first model to optimize early spatio-temporal forecasts using multi-objective reinforcement learning, adapting to preferences or inferring them from limited data.

Findings

Outperforms existing methods on three large-scale datasets

Enhances early forecasting accuracy while maintaining timeliness

Provides an optimal policy for prediction timing in spatial areas

Abstract

Accuracy and timeliness are indeed often conflicting goals in prediction tasks. Premature predictions may yield a higher rate of false alarms, whereas delaying predictions to gather more information can render them too late to be useful. In applications such as wildfires, crimes, and traffic jams, timely forecasting are vital for safeguarding human life and property. Consequently, finding a balance between accuracy and timeliness is crucial. In this paper, we propose an early spatio-temporal forecasting model based on Multi-Objective reinforcement learning that can either implement an optimal policy given a preference or infer the preference based on a small number of samples. The model addresses two primary challenges: 1) enhancing the accuracy of early forecasting and 2) providing the optimal policy for determining the most suitable prediction time for each area. Our method demonstrates superior performance on three large-scale real-world datasets, surpassing existing methods in early spatio-temporal forecasting tasks.