Stochastic Sequential Decision Making over Expanding Networks with Graph Filtering

TL;DR

This paper introduces a novel stochastic sequential decision-making framework for graph filtering on expanding networks, utilizing multi-agent reinforcement learning and context-aware graph neural networks to improve long-term decision quality.

Contribution

It develops a new adaptive filtering approach that accounts for graph expansion and future impacts, unlike existing methods that are limited to fixed graphs or myopic views.

Findings

Outperforms batch and online filtering methods on synthetic and real datasets.

Effectively models graph expansion dynamics through multi-agent reinforcement learning.

Enhances decision-making in applications like recommendation and COVID prediction.

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.