Exploring Self-training for Imbalanced Node Classification

Xin Juan, Meixin Peng, Xin Wang

Abstract: Graph Neural Networks (GNNs) have achieved tremendous success in various applications. However, datasets are usually imbalanced in many real-world scenarios. When trained on an imbalanced dataset, the performance of GNNs is distant from satisfactory for nodes of minority classes. Due to the small population, these minority nodes have less engagement in the objective function of training and the message-passing mechanism behind GNNs exacerbates this problem further. Hence, in this paper, we present a novel model-agnostic framework, named SET-GNN, which utilizes self-training to expand the labeled training set through using pseudo-labeled nodes for improving the performance of the semi-supervised learning based on GNNs for imbalanced node classification. Extensive experiments on various real-world datasets demonstrate SET-GNN is able to achieve state-of-the-art performance on solving the imbalanced node classification problem.