Residual Network and Embedding Usage: New Tricks of Node Classification with Graph Convolutional Networks

TL;DR

This paper introduces new techniques leveraging residual networks and pre-trained embeddings to enhance node classification accuracy in Graph Convolutional Networks, demonstrating significant improvements on benchmark datasets.

Contribution

The paper proposes two novel tricks, GCN_res Framework and Embedding Usage, to improve GCN performance by integrating residual connections and pre-trained embeddings.

Findings

Test accuracy increased by 1.21% to 2.84% on OGB datasets.

Combining techniques yields consistent performance improvements.

Open source implementation provided for reproducibility.

Abstract

Graph Convolutional Networks (GCNs) and subsequent variants have been proposed to solve tasks on graphs, especially node classification tasks. In the literature, however, most tricks or techniques are either briefly mentioned as implementation details or only visible in source code. In this paper, we first summarize some existing effective tricks used in GCNs mini-batch training. Based on this, two novel tricks named GCN_res Framework and Embedding Usage are proposed by leveraging residual network and pre-trained embedding to improve baseline's test accuracy in different datasets. Experiments on Open Graph Benchmark (OGB) show that, by combining these techniques, the test accuracy of various GCNs increases by 1.21%~2.84%. We open source our implementation at https://github.com/ytchx1999/PyG-OGB-Tricks.