A study on Ensemble Learning for Time Series Forecasting and the need for Meta-Learning
This work addresses the challenge of optimizing ensemble strategies for time series forecasting, which is incremental as it builds on existing ensemble techniques by adding a meta-learning layer for customization.
The study tackled the problem of improving time series forecasting by exploring ensemble methods, finding that while ensembles generally enhance results across over 16,000 datasets, no single strategy consistently outperforms others. To address this, they proposed a meta-learning approach to select the best ensemble method and hyperparameters for each dataset based on meta-features.
The contribution of this work is twofold: (1) We introduce a collection of ensemble methods for time series forecasting to combine predictions from base models. We demonstrate insights on the power of ensemble learning for forecasting, showing experiment results on about 16000 openly available datasets, from M4, M5, M3 competitions, as well as FRED (Federal Reserve Economic Data) datasets. Whereas experiments show that ensembles provide a benefit on forecasting results, there is no clear winning ensemble strategy (plus hyperparameter configuration). Thus, in addition, (2), we propose a meta-learning step to choose, for each dataset, the most appropriate ensemble method and their hyperparameter configuration to run based on dataset meta-features.