Non-Recursive Graph Convolutional Networks

TL;DR

This paper introduces NRGCN, a novel GCN architecture that avoids recursive neighborhood aggregation, leading to improved training efficiency and better node classification performance by independently representing neighbor hops.

Contribution

NRGCN proposes a layer-independent sampling and aggregation method that precomputes neighbor information, enhancing efficiency and performance over traditional recursive GCNs.

Findings

NRGCN outperforms state-of-the-art GCNs in node classification accuracy.

Precomputing neighbor information accelerates training significantly.

NRGCN reduces feature redundancy and information loss.

Abstract

Graph Convolutional Networks (GCNs) are powerful models for node representation learning tasks. However, the node representation in existing GCN models is usually generated by performing recursive neighborhood aggregation across multiple graph convolutional layers with certain sampling methods, which may lead to redundant feature mixing, needless information loss, and extensive computations. Therefore, in this paper, we propose a novel architecture named Non-Recursive Graph Convolutional Network (NRGCN) to improve both the training efficiency and the learning performance of GCNs in the context of node classification. Specifically, NRGCN proposes to represent different hops of neighbors for each node based on inner-layer aggregation and layer-independent sampling. In this way, each node can be directly represented by concatenating the information extracted independently from each hop of its neighbors thereby avoiding the recursive neighborhood expansion across layers. Moreover, the layer-independent sampling and aggregation can be precomputed before the model training, thus the training process can be accelerated considerably. Extensive experiments on benchmark datasets verify that our NRGCN outperforms the state-of-the-art GCN models, in terms of the node classification performance and reliability.