Multi-layer Stack Ensembles for Time Series Forecasting
This work addresses the problem of improving accuracy in time series forecasting for AutoML systems, though it is incremental as it builds on existing stacking techniques.
The paper tackled the underutilization of ensemble methods in time series forecasting by systematically evaluating 33 ensemble models across 50 datasets, finding that stacking improves accuracy but no single stacker is best, and proposed a multi-layer stacking framework that consistently provides superior accuracy.
Ensembling is a powerful technique for improving the accuracy of machine learning models, with methods like stacking achieving strong results in tabular tasks. In time series forecasting, however, ensemble methods remain underutilized, with simple linear combinations still considered state-of-the-art. In this paper, we systematically explore ensembling strategies for time series forecasting. We evaluate 33 ensemble models -- both existing and novel -- across 50 real-world datasets. Our results show that stacking consistently improves accuracy, though no single stacker performs best across all tasks. To address this, we propose a multi-layer stacking framework for time series forecasting, an approach that combines the strengths of different stacker models. We demonstrate that this method consistently provides superior accuracy across diverse forecasting scenarios. Our findings highlight the potential of stacking-based methods to improve AutoML systems for time series forecasting.