SwiftTS: A Swift Selection Framework for Time Series Pre-trained Models via Multi-task Meta-Learning
This addresses the time-consuming process of model selection for time series applications, offering a practical solution for researchers and practitioners, though it is incremental as it builds on existing pre-trained models and selection methods.
The paper tackles the problem of efficiently selecting the most suitable pre-trained model for time series tasks by proposing SwiftTS, a framework that predicts model performance using historical data and avoids costly fine-tuning, achieving state-of-the-art results on 14 datasets and 8 models.
Pre-trained models exhibit strong generalization to various downstream tasks. However, given the numerous models available in the model hub, identifying the most suitable one by individually fine-tuning is time-consuming. In this paper, we propose \textbf{SwiftTS}, a swift selection framework for time series pre-trained models. To avoid expensive forward propagation through all candidates, SwiftTS adopts a learning-guided approach that leverages historical dataset-model performance pairs across diverse horizons to predict model performance on unseen datasets. It employs a lightweight dual-encoder architecture that embeds time series and candidate models with rich characteristics, computing patchwise compatibility scores between data and model embeddings for efficient selection. To further enhance the generalization across datasets and horizons, we introduce a horizon-adaptive expert composition module that dynamically adjusts expert weights, and the transferable cross-task learning with cross-dataset and cross-horizon task sampling to enhance out-of-distribution (OOD) robustness. Extensive experiments on 14 downstream datasets and 8 pre-trained models demonstrate that SwiftTS achieves state-of-the-art performance in time series pre-trained model selection.