Infinite-Horizon Graph Filters: Leveraging Power Series to Enhance Sparse Information Aggregation

arXiv:2401.09943v3 Announce Type: replace
Abstract: Graph Neural Networks (GNNs) have shown considerable effectiveness in a variety of graph learning tasks, particularly those based on the message-passing approach in recent years. However, their performance is often constrained by a limited receptive field, a challenge that becomes more acute in the presence of sparse graphs. In light of the power series, which possesses infinite expansion capabilities, we propose a novel Graph Power Filter Neural Network (GPFN) that enhances node classification by employing …

abstract aggregation arxiv challenge cs.lg cs.si filters gnns graph graph learning graph neural networks horizon however information networks neural networks performance power series tasks type