G-Mixup: Graph Data Augmentation for Graph Classification

Xiaotian Han, Zhimeng Jiang, Ninghao Liu, Xia Hu
Proceedings of the 39th International Conference on Machine Learning, PMLR 162:8230-8248, 2022.

Abstract

This work develops mixup for graph data. Mixup has shown superiority in improving the generalization and robustness of neural networks by interpolating features and labels between two random samples. Traditionally, Mixup can work on regular, grid-like, and Euclidean data such as image or tabular data. However, it is challenging to directly adopt Mixup to augment graph data because different graphs typically: 1) have different numbers of nodes; 2) are not readily aligned; and 3) have unique typologies in non-Euclidean space. To this end, we propose G-Mixup to augment graphs for graph classification by interpolating the generator (i.e., graphon) of different classes of graphs. Specifically, we first use graphs within the same class to estimate a graphon. Then, instead of directly manipulating graphs, we interpolate graphons of different classes in the Euclidean space to get mixed graphons, where the synthetic graphs are generated through sampling based on the mixed graphons. Extensive experiments show that G-Mixup substantially improves the generalization and robustness of GNNs.

Cite this Paper

BibTeX
@InProceedings{pmlr-v162-han22c, title = {G-Mixup: Graph Data Augmentation for Graph Classification}, author = {Han, Xiaotian and Jiang, Zhimeng and Liu, Ninghao and Hu, Xia}, booktitle = {Proceedings of the 39th International Conference on Machine Learning}, pages = {8230--8248}, 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/han22c/han22c.pdf}, url = {https://proceedings.mlr.press/v162/han22c.html}, abstract = {This work develops mixup for graph data. Mixup has shown superiority in improving the generalization and robustness of neural networks by interpolating features and labels between two random samples. Traditionally, Mixup can work on regular, grid-like, and Euclidean data such as image or tabular data. However, it is challenging to directly adopt Mixup to augment graph data because different graphs typically: 1) have different numbers of nodes; 2) are not readily aligned; and 3) have unique typologies in non-Euclidean space. To this end, we propose G-Mixup to augment graphs for graph classification by interpolating the generator (i.e., graphon) of different classes of graphs. Specifically, we first use graphs within the same class to estimate a graphon. Then, instead of directly manipulating graphs, we interpolate graphons of different classes in the Euclidean space to get mixed graphons, where the synthetic graphs are generated through sampling based on the mixed graphons. Extensive experiments show that G-Mixup substantially improves the generalization and robustness of GNNs.} }
Endnote
%0 Conference Paper %T G-Mixup: Graph Data Augmentation for Graph Classification %A Xiaotian Han %A Zhimeng Jiang %A Ninghao Liu %A Xia Hu %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-han22c %I PMLR %P 8230--8248 %U https://proceedings.mlr.press/v162/han22c.html %V 162 %X This work develops mixup for graph data. Mixup has shown superiority in improving the generalization and robustness of neural networks by interpolating features and labels between two random samples. Traditionally, Mixup can work on regular, grid-like, and Euclidean data such as image or tabular data. However, it is challenging to directly adopt Mixup to augment graph data because different graphs typically: 1) have different numbers of nodes; 2) are not readily aligned; and 3) have unique typologies in non-Euclidean space. To this end, we propose G-Mixup to augment graphs for graph classification by interpolating the generator (i.e., graphon) of different classes of graphs. Specifically, we first use graphs within the same class to estimate a graphon. Then, instead of directly manipulating graphs, we interpolate graphons of different classes in the Euclidean space to get mixed graphons, where the synthetic graphs are generated through sampling based on the mixed graphons. Extensive experiments show that G-Mixup substantially improves the generalization and robustness of GNNs.
APA
Han, X., Jiang, Z., Liu, N. & Hu, X.. (2022). G-Mixup: Graph Data Augmentation for Graph Classification. Proceedings of the 39th International Conference on Machine Learning, in Proceedings of Machine Learning Research 162:8230-8248 Available from https://proceedings.mlr.press/v162/han22c.html.

Related Material