Stochastic Sequential Decision Making over Expanding Networks with Graph Filtering

2026-03-19

AI Summary

This work addresses the limitations of existing graph filtering methods, which typically assume a static graph structure and struggle to handle continuous, pattern-agnostic node expansion while lacking a decision mechanism that jointly considers historical and future influences. To overcome these challenges, we propose a stochastic sequential decision-making framework tailored for expanding networks, introducing multi-agent reinforcement learning into graph filtering for the first time. In our approach, filter shifts are modeled as a multi-agent system, and context-aware graph neural networks dynamically adjust filtering parameters. Extensive experiments on both real-world and synthetic datasets—including cold-start recommendation and COVID-19 forecasting—demonstrate that our method significantly outperforms state-of-the-art batch and online filtering approaches, thereby validating the efficacy and superiority of a sequential decision-making perspective in dynamic graph filtering.

Technology Category

• Machine Learning: Graph-based Machine Learning
• Reasoning under Uncertainty: Sequential Decision Making
• Data Mining & Knowledge Management: Graph Mining, Social Network Analysis & Community

Application Category

• Graph Algorithms and Modeling for the Web: Efficient manipulation of static and dynamic Web-related graphs
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• User Modeling, Personalization and Recommendation: Attacks and countermeasures in recommendation systems

Abstract

Graph filters leverage topological information to process networked data with existing methods mainly studying fixed graphs, ignoring that graphs often expand as nodes continually attach with an unknown pattern. The latter requires developing filter-based decision-making paradigms that take evolution and uncertainty into account. Existing approaches rely on either pre-designed filters or online learning, limited to a myopic view considering only past or present information. To account for future impacts, we propose a stochastic sequential decision-making framework for filtering networked data with a policy that adapts filtering to expanding graphs. By representing filter shifts as agents, we model the filter as a multi-agent system and train the policy following multi-agent reinforcement learning. This accounts for long-term rewards and captures expansion dynamics through sequential decision-making. Moreover, we develop a context-aware graph neural network to parameterize the policy, which tunes filter parameters based on information of both the graph and agents. Experiments on synthetic and real datasets from cold-start recommendation to COVID prediction highlight the benefits of using a sequential decision-making perspective over batch and online filtering alternatives.

Problem

• expanding networks
• graph filtering
• stochastic sequential decision making
• graph evolution
• uncertainty

Innovation

• stochastic sequential decision making
• expanding graphs
• graph filtering
• multi-agent reinforcement learning
• context-aware graph neural network