CombinationTS: A Modular Framework for Understanding Time-Series Forecasting Models
For researchers in time-series forecasting, this work provides a principled method to attribute performance gains to specific architectural components, challenging the necessity of complex backbones.
The paper introduces CombinationTS, a modular evaluation framework for time-series forecasting that decomposes models into orthogonal components. Using this framework, they discover the Identity Paradox: with a good embedding, a simple Identity Encoder often matches or outperforms complex backbones, and explicit structural priors via Input Transformations yield better performance-stability trade-offs than increasing encoder complexity.
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.