DeepBooTS: Dual-Stream Residual Boosting for Drift-Resilient Time-Series Forecasting
Fuente:
arXiv
Enregistré dans:
| Auteurs principaux: | , , , , |
|---|---|
| Format: | Preprint |
| Publié: |
2025
|
| Sujets: | |
| Accès en ligne: | |
| Tags: |
Ajouter un tag
Pas de tags, Soyez le premier à ajouter un tag!
|
| _version_ | 1866915758367309824 |
|---|---|
| author | Liang, Daojun Chen, Jing Wang, Xiao Wang, Yinglong Li, Shuo |
| author_facet | Liang, Daojun Chen, Jing Wang, Xiao Wang, Yinglong Li, Shuo |
| contents | Time-Series (TS) exhibits pronounced non-stationarity. Consequently, most forecasting methods display compromised robustness to concept drift, despite the prevalent application of instance normalization. We tackle this challenge by first analysing concept drift through a bias-variance lens and proving that weighted ensemble reduces variance without increasing bias. These insights motivate DeepBooTS, a novel end-to-end dual-stream residual-decreasing boosting method that progressively reconstructs the intrinsic signal. In our design, each block of a deep model becomes an ensemble of learners with an auxiliary output branch forming a highway to the final prediction. The block-wise outputs correct the residuals of previous blocks, leading to a learning-driven decomposition of both inputs and targets. This method enhances versatility and interpretability while substantially improving robustness to concept drift. Extensive experiments, including those on large-scale datasets, show that the proposed method outperforms existing methods by a large margin, yielding an average performance improvement of 15.8% across various datasets, establishing a new benchmark for TS forecasting. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2511_06893 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | DeepBooTS: Dual-Stream Residual Boosting for Drift-Resilient Time-Series Forecasting Liang, Daojun Chen, Jing Wang, Xiao Wang, Yinglong Li, Shuo Machine Learning Artificial Intelligence Time-Series (TS) exhibits pronounced non-stationarity. Consequently, most forecasting methods display compromised robustness to concept drift, despite the prevalent application of instance normalization. We tackle this challenge by first analysing concept drift through a bias-variance lens and proving that weighted ensemble reduces variance without increasing bias. These insights motivate DeepBooTS, a novel end-to-end dual-stream residual-decreasing boosting method that progressively reconstructs the intrinsic signal. In our design, each block of a deep model becomes an ensemble of learners with an auxiliary output branch forming a highway to the final prediction. The block-wise outputs correct the residuals of previous blocks, leading to a learning-driven decomposition of both inputs and targets. This method enhances versatility and interpretability while substantially improving robustness to concept drift. Extensive experiments, including those on large-scale datasets, show that the proposed method outperforms existing methods by a large margin, yielding an average performance improvement of 15.8% across various datasets, establishing a new benchmark for TS forecasting. |
| title | DeepBooTS: Dual-Stream Residual Boosting for Drift-Resilient Time-Series Forecasting |
| topic | Machine Learning Artificial Intelligence |
| url | https://arxiv.org/abs/2511.06893 |