Calibrating and Improving Graph Contrastive Learning

Abstract

Graph contrastive learning algorithms have demonstrated remarkable success in various applications such as node classification, link prediction, and graph clustering. However, in unsupervised graph contrastive learning, some contrastive pairs may contradict the truths in downstream tasks and thus the decrease of losses on these pairs undesirably harms the performance in the downstream tasks. To assess the discrepancy between the prediction and the ground-truth in the downstream tasks for these contrastive pairs, we adapt the expected calibration error (ECE) to graph contrastive learning. The analysis of ECE motivates us to propose a novel regularization method, Contrast-Reg, to ensure that decreasing the contrastive loss leads to better performance in the downstream tasks. As a plug-in regularizer, Contrast-Reg effectively improves the performance of existing graph contrastive learning algorithms. We provide both theoretical and empirical results to demonstrate the effectiveness of Contrast-Reg in enhancing the generalizability of the Graph Neural Network(GNN) model and improving the performance of graph contrastive algorithms with different similarity definitions and encoder backbones across various downstream tasks.

References in corpus (7)

• Modularity and community structure in networks
• Fast Graph Representation Learning with PyTorch Geometric
• A Theoretical Analysis of Contrastive Unsupervised Representation Learning
• HARP: Hierarchical Representation Learning for Networks
• Unsupervised Deep Learning by Neighbourhood Discovery
• Notes on Noise Contrastive Estimation and Negative Sampling
• Batch Virtual Adversarial Training for Graph Convolutional Networks

Cited by in corpus (3)

• Graph Self-Supervised Learning: A Survey
• AKE-GNN: Effective Graph Learning with Adaptive Knowledge Exchange
• Barlow Graph Auto-Encoder for Unsupervised Network Embedding