Saved in:
| Main Authors: | , , , , , , , , , , , |
|---|---|
| Format: | Preprint |
| Published: |
2026
|
| Subjects: | |
| Online Access: | https://arxiv.org/abs/2605.01231 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866913083444690944 |
|---|---|
| author | Wang, Xiaorui Fan, Fanda Wang, Chenxi Yang, Yuxuan Tang, Rui Gao, Kuoyu Pang, Simiao Shang, Yuanfeng Liu, Zhipeng Gao, Wanling Wang, Lei Zhan, Jianfeng |
| author_facet | Wang, Xiaorui Fan, Fanda Wang, Chenxi Yang, Yuxuan Tang, Rui Gao, Kuoyu Pang, Simiao Shang, Yuanfeng Liu, Zhipeng Gao, Wanling Wang, Lei Zhan, Jianfeng |
| contents | Recent progress in time-series forecasting has led to rapidly increasing architectural complexity, yet many reported State-of-the-Art gains are statistically fragile or misattributed. We argue that progress requires a shift from model selection to modular attribution, identifying which components truly drive performance. We propose CombinationTS, a self-contained probabilistic evaluation framework that decomposes forecasting models into orthogonal modules--Input Transformation, Embedding, Encoder, Decoder, and Output Transformation--and evaluates them under a shared evaluation condition space. By quantifying each component via marginalized performance ($μ$) and stability ($σ$), CombinationTS enables robust attribution beyond fragile point estimates. Through large-scale paired evaluation, we uncover the Identity Paradox: once the data view (Embedding) is well-designed, a parameter-free Identity Encoder often matches or outperforms complex backbones. We further show that explicit structural priors introduced via Input Transformations yield a more favorable performance-stability trade-off than increasing Encoder complexity, establishing a principled baseline for architectural necessity. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2605_01231 |
| institution | arXiv |
| publishDate | 2026 |
| record_format | arxiv |
| spellingShingle | CombinationTS: A Modular Framework for Understanding Time-Series Forecasting Models Wang, Xiaorui Fan, Fanda Wang, Chenxi Yang, Yuxuan Tang, Rui Gao, Kuoyu Pang, Simiao Shang, Yuanfeng Liu, Zhipeng Gao, Wanling Wang, Lei Zhan, Jianfeng Machine Learning Recent progress in time-series forecasting has led to rapidly increasing architectural complexity, yet many reported State-of-the-Art gains are statistically fragile or misattributed. We argue that progress requires a shift from model selection to modular attribution, identifying which components truly drive performance. We propose CombinationTS, a self-contained probabilistic evaluation framework that decomposes forecasting models into orthogonal modules--Input Transformation, Embedding, Encoder, Decoder, and Output Transformation--and evaluates them under a shared evaluation condition space. By quantifying each component via marginalized performance ($μ$) and stability ($σ$), CombinationTS enables robust attribution beyond fragile point estimates. Through large-scale paired evaluation, we uncover the Identity Paradox: once the data view (Embedding) is well-designed, a parameter-free Identity Encoder often matches or outperforms complex backbones. We further show that explicit structural priors introduced via Input Transformations yield a more favorable performance-stability trade-off than increasing Encoder complexity, establishing a principled baseline for architectural necessity. |
| title | CombinationTS: A Modular Framework for Understanding Time-Series Forecasting Models |
| topic | Machine Learning |
| url | https://arxiv.org/abs/2605.01231 |