Inductive Spatial Temporal Prediction Under Data Drift with Informative Graph Neural Network

TL;DR

This paper introduces INF-GNN, a graph neural network designed to improve inductive spatial-temporal predictions under data drift by capturing diversified invariant patterns and emphasizing influential temporal information.

Contribution

The paper proposes a novel INF-GNN model with an informative subgraph and temporal memory buffer to enhance generalization to unseen data under data drift conditions.

Findings

INF-GNN outperforms existing methods on real-world datasets.

The informative subgraph effectively selects stable entities and relationships.

Temporal memory improves the recognition of influential time patterns.

Abstract

Inductive spatial temporal prediction can generalize historical data to predict unseen data, crucial for highly dynamic scenarios (e.g., traffic systems, stock markets). However, external events (e.g., urban structural growth, market crash) and emerging new entities (e.g., locations, stocks) can undermine prediction accuracy by inducing data drift over time. Most existing studies extract invariant patterns to counter data drift but ignore pattern diversity, exhibiting poor generalization to unseen entities. To address this issue, we design an Informative Graph Neural Network (INF-GNN) to distill diversified invariant patterns and improve prediction accuracy under data drift. Firstly, we build an informative subgraph with a uniquely designed metric, Relation Importance (RI), that can effectively select stable entities and distinct spatial relationships. This subgraph further generalizes new entities' data via neighbors merging. Secondly, we propose an informative temporal memory buffer to help the model emphasize valuable timestamps extracted using influence functions within time intervals. This memory buffer allows INF-GNN to discern influential temporal patterns. Finally, RI loss optimization is designed for pattern consolidation. Extensive experiments on real-world dataset under substantial data drift demonstrate that INF-GNN significantly outperforms existing alternatives.