Graph Neural Architecture Search Under Distribution Shifts

Yijian Qin, Xin Wang, Ziwei Zhang, Pengtao Xie, Wenwu Zhu
Proceedings of the 39th International Conference on Machine Learning, PMLR 162:18083-18095, 2022.

Abstract

Graph neural architecture search has shown great potentials for automatically designing graph neural network (GNN) architectures for graph classification tasks. However, when there is a distribution shift between training and testing graphs, the existing approaches fail to deal with the problem of adapting to unknown test graph structures since they only search for a fixed architecture for all graphs. To solve this problem, we propose a novel GRACES model which is able to generalize under distribution shifts through tailoring a customized GNN architecture suitable for each graph instance with unknown distribution. Specifically, we design a self-supervised disentangled graph encoder to characterize invariant factors hidden in diverse graph structures. Then, we propose a prototype-based architecture customization strategy to generate the most suitable GNN architecture weights in a continuous space for each graph instance. We further propose a customized super-network to share weights among different architectures for the sake of efficient training. Extensive experiments on both synthetic and real-world datasets demonstrate that our proposed GRACES model can adapt to diverse graph structures and achieve state-of-the-art performance for graph classification tasks under distribution shifts.

Cite this Paper


BibTeX
@InProceedings{pmlr-v162-qin22b, title = {Graph Neural Architecture Search Under Distribution Shifts}, author = {Qin, Yijian and Wang, Xin and Zhang, Ziwei and Xie, Pengtao and Zhu, Wenwu}, booktitle = {Proceedings of the 39th International Conference on Machine Learning}, pages = {18083--18095}, 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/qin22b/qin22b.pdf}, url = {https://proceedings.mlr.press/v162/qin22b.html}, abstract = {Graph neural architecture search has shown great potentials for automatically designing graph neural network (GNN) architectures for graph classification tasks. However, when there is a distribution shift between training and testing graphs, the existing approaches fail to deal with the problem of adapting to unknown test graph structures since they only search for a fixed architecture for all graphs. To solve this problem, we propose a novel GRACES model which is able to generalize under distribution shifts through tailoring a customized GNN architecture suitable for each graph instance with unknown distribution. Specifically, we design a self-supervised disentangled graph encoder to characterize invariant factors hidden in diverse graph structures. Then, we propose a prototype-based architecture customization strategy to generate the most suitable GNN architecture weights in a continuous space for each graph instance. We further propose a customized super-network to share weights among different architectures for the sake of efficient training. Extensive experiments on both synthetic and real-world datasets demonstrate that our proposed GRACES model can adapt to diverse graph structures and achieve state-of-the-art performance for graph classification tasks under distribution shifts.} }
Endnote
%0 Conference Paper %T Graph Neural Architecture Search Under Distribution Shifts %A Yijian Qin %A Xin Wang %A Ziwei Zhang %A Pengtao Xie %A Wenwu Zhu %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-qin22b %I PMLR %P 18083--18095 %U https://proceedings.mlr.press/v162/qin22b.html %V 162 %X Graph neural architecture search has shown great potentials for automatically designing graph neural network (GNN) architectures for graph classification tasks. However, when there is a distribution shift between training and testing graphs, the existing approaches fail to deal with the problem of adapting to unknown test graph structures since they only search for a fixed architecture for all graphs. To solve this problem, we propose a novel GRACES model which is able to generalize under distribution shifts through tailoring a customized GNN architecture suitable for each graph instance with unknown distribution. Specifically, we design a self-supervised disentangled graph encoder to characterize invariant factors hidden in diverse graph structures. Then, we propose a prototype-based architecture customization strategy to generate the most suitable GNN architecture weights in a continuous space for each graph instance. We further propose a customized super-network to share weights among different architectures for the sake of efficient training. Extensive experiments on both synthetic and real-world datasets demonstrate that our proposed GRACES model can adapt to diverse graph structures and achieve state-of-the-art performance for graph classification tasks under distribution shifts.
APA
Qin, Y., Wang, X., Zhang, Z., Xie, P. & Zhu, W.. (2022). Graph Neural Architecture Search Under Distribution Shifts. Proceedings of the 39th International Conference on Machine Learning, in Proceedings of Machine Learning Research 162:18083-18095 Available from https://proceedings.mlr.press/v162/qin22b.html.

Related Material