Track papers across labs, topics, and trends.
Search papers, labs, and topics across Lattice.
This paper introduces TSCOMP, a large-scale benchmark for multivariate time series forecasting that decomposes deep forecasting methods into fine-grained components like preprocessing, encoding, architectures, and optimization. Through constrained orthogonal experimental design and over 20,000 model-dataset evaluations, the authors analyze component effectiveness across different data characteristics and model backbones. They demonstrate that a corpus-driven approach to component selection outperforms state-of-the-art holistic methods, enabling zero-shot model construction on new datasets.
Forget complex architectures: zero-shot time-series forecasting is now possible by intelligently composing simpler, pre-evaluated components.
While previous research in multivariate time series forecasting has focused on developing complex holistic models, this work advocates for a shift toward a granular, component-level understanding of their impacts. We propose TSCOMP, the first large-scale benchmark that systematically deconstructs deep forecasting methods into their core, fine-grained components--spanning series preprocessing, encoding strategies, network architectures including specific and large time-series models, and optimization methods. Using constrained orthogonal experimental design and extensive evaluations, we conduct multi-view analyses that reveal component effectiveness across different backbones, data characteristics, and their interactions. Beyond providing insights, this benchmark establishes a fine-grained performance corpus comprising over 20,000 model-dataset evaluations, which supports the learning of automated component selection, enabling zero-shot model construction on new datasets. Our experiments demonstrate that the corpus-driven approach, despite its simplicity, consistently outperforms state-of-the-art methods, validating the soundness of our evaluation design and confirming that systematic component selection surpasses manually designed complex architectures. All code and the performance corpus are publicly available at https://github.com/SUFE-AILAB/TSCOMP.