TAM: Topology-Aware Margin Loss for Class-Imbalanced Node Classification

Jaeyun Song, Joonhyung Park, Eunho Yang
Proceedings of the 39th International Conference on Machine Learning, PMLR 162:20369-20383, 2022.

Abstract

Learning unbiased node representations under class-imbalanced graph data is challenging due to interactions between adjacent nodes. Existing studies have in common that they compensate the minor class nodes ‘as a group’ according to their overall quantity (ignoring node connections in graph), which inevitably increase the false positive cases for major nodes. We hypothesize that the increase in these false positive cases is highly affected by the label distribution around each node and confirm it experimentally. In addition, in order to handle this issue, we propose Topology-Aware Margin (TAM) to reflect local topology on the learning objective. Our method compares the connectivity pattern of each node with the class-averaged counter-part and adaptively adjusts the margin accordingly based on that. Our method consistently exhibits superiority over the baselines on various node classification benchmark datasets with representative GNN architectures.

Cite this Paper

BibTeX
@InProceedings{pmlr-v162-song22a, title = {{TAM}: Topology-Aware Margin Loss for Class-Imbalanced Node Classification}, author = {Song, Jaeyun and Park, Joonhyung and Yang, Eunho}, booktitle = {Proceedings of the 39th International Conference on Machine Learning}, pages = {20369--20383}, year = {2022}, editor = {Chaudhuri, Kamalika and Jegelka, Stefanie and Song, Le and Szepesvari, Csaba and Niu, Gang and Sabato, Sivan}, volume = {162}, series = {Proceedings of Machine Learning Research}, month = {17--23 Jul}, publisher = {PMLR}, pdf = {https://proceedings.mlr.press/v162/song22a/song22a.pdf}, url = {https://proceedings.mlr.press/v162/song22a.html}, abstract = {Learning unbiased node representations under class-imbalanced graph data is challenging due to interactions between adjacent nodes. Existing studies have in common that they compensate the minor class nodes ‘as a group’ according to their overall quantity (ignoring node connections in graph), which inevitably increase the false positive cases for major nodes. We hypothesize that the increase in these false positive cases is highly affected by the label distribution around each node and confirm it experimentally. In addition, in order to handle this issue, we propose Topology-Aware Margin (TAM) to reflect local topology on the learning objective. Our method compares the connectivity pattern of each node with the class-averaged counter-part and adaptively adjusts the margin accordingly based on that. Our method consistently exhibits superiority over the baselines on various node classification benchmark datasets with representative GNN architectures.} }
Endnote
%0 Conference Paper %T TAM: Topology-Aware Margin Loss for Class-Imbalanced Node Classification %A Jaeyun Song %A Joonhyung Park %A Eunho Yang %B Proceedings of the 39th International Conference on Machine Learning %C Proceedings of Machine Learning Research %D 2022 %E Kamalika Chaudhuri %E Stefanie Jegelka %E Le Song %E Csaba Szepesvari %E Gang Niu %E Sivan Sabato %F pmlr-v162-song22a %I PMLR %P 20369--20383 %U https://proceedings.mlr.press/v162/song22a.html %V 162 %X Learning unbiased node representations under class-imbalanced graph data is challenging due to interactions between adjacent nodes. Existing studies have in common that they compensate the minor class nodes ‘as a group’ according to their overall quantity (ignoring node connections in graph), which inevitably increase the false positive cases for major nodes. We hypothesize that the increase in these false positive cases is highly affected by the label distribution around each node and confirm it experimentally. In addition, in order to handle this issue, we propose Topology-Aware Margin (TAM) to reflect local topology on the learning objective. Our method compares the connectivity pattern of each node with the class-averaged counter-part and adaptively adjusts the margin accordingly based on that. Our method consistently exhibits superiority over the baselines on various node classification benchmark datasets with representative GNN architectures.
APA
Song, J., Park, J. & Yang, E.. (2022). TAM: Topology-Aware Margin Loss for Class-Imbalanced Node Classification. Proceedings of the 39th International Conference on Machine Learning, in Proceedings of Machine Learning Research 162:20369-20383 Available from https://proceedings.mlr.press/v162/song22a.html.

Related Material