OMG: Towards Effective Graph Classification Against Label Noise

Document Type

Article

Publication Title

IEEE Transactions on Knowledge and Data Engineering

Abstract

Graph classification is a fundamental problem with diverse applications in bioinformatics and chemistry. Due to the intricate procedures of manual annotations in graphical domains, there may be abundant noisy labels of graphs in practice, resulting in poor performance for existing supervised methods. Thus, it is necessary and urgent to study the problem of graph classification with label noise. However, this problem is challenging due to the overfitting of noisy data as well as complicated relational structures of graphs. To handle this problem, we present a simple but effective approach called cOupled Mix for Graph Contrast (OMG), which combines coupled Mixup with graph contrastive learning in the feature space. On the one hand, to improve the model generalization, we take convex combination of sample pairs in the feature space for positive pair construction. On the other hand, to accomplish effective optimization, we offer challenging negatives by multiple sample Mixup with different emphasis. To further reduce the impact of noisy data, we develop a neighbour-aware noise removal strategy, which promotes the smoothness in the neighbourhood of samples following the principle of curriculum learning. Extensive experiments on a range of benchmark datasets demonstrate the superiority of our proposed OMG.

First Page

12873

Last Page

12886

DOI

10.1109/TKDE.2023.3271677

Publication Date

5-1-2023

Keywords

Noise measurement, Training, Kernel, Task analysis, Reliability, Optimization, Graph neural networks

Comments

IR conditions: non-described

Link to Full Text

Share

COinS