Saved in:
Bibliographic Details
Main Authors: Jing, Zihao, Long, Yuxi, Feng, Ganlin
Format: Preprint
Published: 2026
Subjects:
Online Access:https://arxiv.org/abs/2602.04153
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866918322054889472
author Jing, Zihao
Long, Yuxi
Feng, Ganlin
author_facet Jing, Zihao
Long, Yuxi
Feng, Ganlin
contents Multivariate time series forecasting in graph-structured domains is critical for real-world applications, yet existing spatiotemporal models often suffer from performance degradation under data scarcity and cross-domain shifts. We address these challenges through the lens of structure-aware context selection. We propose TL-GPSTGN, a transfer-oriented spatiotemporal framework that enhances sample efficiency and out-of-distribution generalization by selectively pruning non-optimized graph context. Specifically, our method employs information-theoretic and correlation-based criteria to extract structurally informative subgraphs and features, resulting in a compact, semantically grounded representation. This optimized context is subsequently integrated into a spatiotemporal convolutional architecture to capture complex multivariate dynamics. Evaluations on large-scale traffic benchmarks demonstrate that TL-GPSTGN consistently outperforms baselines in low-data transfer scenarios. Our findings suggest that explicit context pruning serves as a powerful inductive bias for improving the robustness of graph-based forecasting models.
format Preprint
id arxiv_https___arxiv_org_abs_2602_04153
institution arXiv
publishDate 2026
record_format arxiv
spellingShingle Pruning for Generalization: A Transfer-Oriented Spatiotemporal Graph Framework
Jing, Zihao
Long, Yuxi
Feng, Ganlin
Machine Learning
Artificial Intelligence
Multivariate time series forecasting in graph-structured domains is critical for real-world applications, yet existing spatiotemporal models often suffer from performance degradation under data scarcity and cross-domain shifts. We address these challenges through the lens of structure-aware context selection. We propose TL-GPSTGN, a transfer-oriented spatiotemporal framework that enhances sample efficiency and out-of-distribution generalization by selectively pruning non-optimized graph context. Specifically, our method employs information-theoretic and correlation-based criteria to extract structurally informative subgraphs and features, resulting in a compact, semantically grounded representation. This optimized context is subsequently integrated into a spatiotemporal convolutional architecture to capture complex multivariate dynamics. Evaluations on large-scale traffic benchmarks demonstrate that TL-GPSTGN consistently outperforms baselines in low-data transfer scenarios. Our findings suggest that explicit context pruning serves as a powerful inductive bias for improving the robustness of graph-based forecasting models.
title Pruning for Generalization: A Transfer-Oriented Spatiotemporal Graph Framework
topic Machine Learning
Artificial Intelligence
url https://arxiv.org/abs/2602.04153