Towards Semi-Supervised Universal Graph Classification
Top Cited Papers
- 29 May 2023
- journal article
- research article
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Knowledge and Data Engineering
- Vol. 36 (1) , 416-428
Abstract
Graph neural networks have pushed state-of-the-arts in graph classifications recently. Typically, these methods are studied within the context of supervised end-to-end training, which necessities copious task-specific labels. However, in real-world circumstances, labeled data could be limited, and there could be a massive corpus of unlabeled data, even from unknown classes as a complementary. Towards this end, we study the problem of semi-supervised universal graph classification, which not only identifies graph samples which do not belong to known classes, but also classifies the remaining samples into their respective classes. This problem is challenging due to a severe lack of labels and potential class shifts. In this paper, we propose a novel graph neural network framework named UGNN, which makes the best of unlabeled data from the subgraph perspective. To tackle class shifts, we estimate the certainty of unlabeled graphs using multiple subgraphs, which facilities the discovery of unlabeled data from unknown categories. Moreover, we construct semantic prototypes in the embedding space for both known and unknown categories and utilize posterior prototype assignments inferred from the Sinkhorn-Knopp algorithm to learn from abundant unlabeled graphs across different subgraph views. Extensive experiments on six datasets verify the effectiveness of UGNN in different settings.Keywords
All Related Versions
Funding Information
- National Natural Science Foundation of China (62276002)
This publication has 26 references indexed in Scilit:
- ASAP: Adaptive Structure Aware Pooling for Learning Hierarchical Graph RepresentationsProceedings of the AAAI Conference on Artificial Intelligence, 2020
- Deep Learning on Graphs: A SurveyIEEE Transactions on Knowledge and Data Engineering, 2020
- Molecular Property Prediction: A Multilevel Quantum Interactions Modeling PerspectiveProceedings of the AAAI Conference on Artificial Intelligence, 2019
- Why ReLU Networks Yield High-Confidence Predictions Far Away From the Training Data and How to Mitigate the Problem2019
- Semi-Supervised Graph Classification: A Hierarchical Graph Perspective2019
- Out-of-Distribution Detection Using an Ensemble of Self Supervised Leave-Out Classifiers2018
- Graph Convolutional Neural Networks for Web-Scale Recommender Systems2018
- Machine Learning Predictions of Molecular Properties: Accurate Many-Body Potentials and Nonlocality in Chemical SpaceThe Journal of Physical Chemistry Letters, 2015
- IAM Graph Database Repository for Graph Based Pattern Recognition and Machine Learning2008
- Shortest-Path Kernels on Graphs2005