GraphTTA: Test Time Adaptation on Graph Neural Networks

TL;DR

This paper introduces GraphTTA, a novel test time adaptation method for GNNs called GAPGC, which uses contrastive learning with adversarial augmentation and pseudo positives to improve OOD performance, backed by theoretical analysis and experiments.

Contribution

The paper proposes a new TTA strategy for GNNs, combining contrastive learning with adversarial augmentation and pseudo positives, tailored for irregular graph structures.

Findings

Achieves state-of-the-art OOD performance on molecular datasets.

Theoretically demonstrates minimal sufficient information extraction.

Effective adaptation for graph neural networks under distribution shifts.

Abstract

Recently, test time adaptation (TTA) has attracted increasing attention due to its power of handling the distribution shift issue in the real world. Unlike what has been developed for convolutional neural networks (CNNs) for image data, TTA is less explored for Graph Neural Networks (GNNs). There is still a lack of efficient algorithms tailored for graphs with irregular structures. In this paper, we present a novel test time adaptation strategy named Graph Adversarial Pseudo Group Contrast (GAPGC), for graph neural networks TTA, to better adapt to the Out Of Distribution (OOD) test data. Specifically, GAPGC employs a contrastive learning variant as a self-supervised task during TTA, equipped with Adversarial Learnable Augmenter and Group Pseudo-Positive Samples to enhance the relevance between the self-supervised task and the main task, boosting the performance of the main task. Furthermore, we provide theoretical evidence that GAPGC can extract minimal sufficient information for the main task from information theory perspective. Extensive experiments on molecular scaffold OOD dataset demonstrated that the proposed approach achieves state-of-the-art performance on GNNs.