Reproduce “T2ST: Data-Efficient Spatio-Temporal Foundation Models via Temporal Data and Knowledge Transfer” on OpenResearch
Turn your favorite coding agents into research agents that review literature, develop hypotheses, and run experiments.
Universal app for Apple silicon and Intel. Requires macOS 11 or later.
About this paper
Yue Xu, Kaiyun Wei, Wenying Duan, Zichen Xu, Shuai Ma, Xiaoxi He
T2ST uses selected external time series and frozen TimesFM 2.5 to strengthen temporal learning, while real spatio-temporal observations ground spatial learning. It selects matched and diverse series and transfers temporal relations and multi-scale forecasts to a student. Using 10% of the structured data, it achieves competitive or better forecasting across six long-term and twelve short-term datasets, but not uniformly.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.