Detecting the Future: All-at-Once Event Sequence Forecasting with Horizon Matching
This work addresses a crucial task in domains like retail and healthcare by providing a more accurate and efficient forecasting method, though it appears incremental as it builds on existing event prediction models.
The paper tackled the problem of long-horizon event sequence forecasting, which often suffers from limited accuracy and diversity due to autoregressive methods, by introducing DEF, a novel approach that simultaneously predicts multiple future events with a matching-based loss function, achieving up to a 50% relative improvement over existing models.
Long-horizon events forecasting is a crucial task across various domains, including retail, finance, healthcare, and social networks. Traditional models for event sequences often extend to forecasting on a horizon using an autoregressive (recursive) multi-step strategy, which has limited effectiveness due to typical convergence to constant or repetitive outputs. To address this limitation, we introduce DEF, a novel approach for simultaneous forecasting of multiple future events on a horizon with high accuracy and diversity. Our method optimally aligns predictions with ground truth events during training by using a novel matching-based loss function. We establish a new state-of-the-art in long-horizon event prediction, achieving up to a 50% relative improvement over existing temporal point processes and event prediction models. Furthermore, we achieve state-of-the-art performance in next-event prediction tasks while demonstrating high computational efficiency during inference.