{"id":22229792,"url":"https://github.com/graph-com/bayesian_inference_based_gnn","last_synced_at":"2025-10-13T08:30:33.111Z","repository":{"id":130189017,"uuid":"549916463","full_name":"Graph-COM/Bayesian_inference_based_GNN","owner":"Graph-COM","description":"[Neurips2022] Understanding Non-linearity in Graph Neural Networks from the Perspective of Bayesian Inference","archived":false,"fork":false,"pushed_at":"2022-10-12T04:10:03.000Z","size":21,"stargazers_count":7,"open_issues_count":0,"forks_count":1,"subscribers_count":0,"default_branch":"main","last_synced_at":"2025-04-06T05:51:12.346Z","etag":null,"topics":[],"latest_commit_sha":null,"homepage":"","language":"Jupyter Notebook","has_issues":true,"has_wiki":null,"has_pages":null,"mirror_url":null,"source_name":null,"license":null,"status":null,"scm":"git","pull_requests_enabled":true,"icon_url":"https://github.com/Graph-COM.png","metadata":{"files":{"readme":"README.md","changelog":null,"contributing":null,"funding":null,"license":null,"code_of_conduct":null,"threat_model":null,"audit":null,"citation":null,"codeowners":null,"security":null,"support":null,"governance":null,"roadmap":null,"authors":null,"dei":null,"publiccode":null,"codemeta":null}},"created_at":"2022-10-11T23:52:28.000Z","updated_at":"2024-03-16T13:48:38.000Z","dependencies_parsed_at":null,"dependency_job_id":"7f2083f7-f66d-4d35-8da9-78d24cf915bc","html_url":"https://github.com/Graph-COM/Bayesian_inference_based_GNN","commit_stats":null,"previous_names":[],"tags_count":0,"template":false,"template_full_name":null,"purl":"pkg:github/Graph-COM/Bayesian_inference_based_GNN","repository_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/Graph-COM%2FBayesian_inference_based_GNN","tags_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/Graph-COM%2FBayesian_inference_based_GNN/tags","releases_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/Graph-COM%2FBayesian_inference_based_GNN/releases","manifests_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/Graph-COM%2FBayesian_inference_based_GNN/manifests","owner_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners/Graph-COM","download_url":"https://codeload.github.com/Graph-COM/Bayesian_inference_based_GNN/tar.gz/refs/heads/main","sbom_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/Graph-COM%2FBayesian_inference_based_GNN/sbom","scorecard":null,"host":{"name":"GitHub","url":"https://github.com","kind":"github","repositories_count":279014326,"owners_count":26085492,"icon_url":"https://github.com/github.png","version":null,"created_at":"2022-05-30T11:31:42.601Z","updated_at":"2022-07-04T15:15:14.044Z","status":"online","status_checked_at":"2025-10-13T02:00:06.723Z","response_time":61,"last_error":null,"robots_txt_status":"success","robots_txt_updated_at":"2025-07-24T06:49:26.215Z","robots_txt_url":"https://github.com/robots.txt","online":true,"can_crawl_api":true,"host_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub","repositories_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories","repository_names_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repository_names","owners_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners"}},"keywords":[],"created_at":"2024-12-03T01:12:26.544Z","updated_at":"2025-10-13T08:30:32.845Z","avatar_url":"https://github.com/Graph-COM.png","language":"Jupyter Notebook","funding_links":[],"categories":[],"sub_categories":[],"readme":"# NeurIPS2022 \"Understanding Non-linearity in Graph Neural Networks from the Bayesian-Inference Perspective\"\nBy Rongzhe Wei, Haoteng Yin, Junteng Jia, Austin R. Benson, Pan Li\n\n## Introduction\nGraph neural networks (GNNs) have become the de-facto standard used in many graph learning tasks due to their super empirical performance. Researchers often attribute such success to non-linearity in GNNs which associates them with great expressive power. However, for node classification tasks, many studies have shown that non-linearity to control the exchange of features among neighbors seems not that crucial. In this work, we resort to understand the effect of non-linearity by comparing with the linear counterparts for node classification tasks from a Bayesian Inference perspective.\n\n## Main Results\n* When the node attributes are less informative compared to the structural information, non-linear\npropagation and linear propagation have almost the same mis-classification error.\n* When the node attributes are more informative, non-linear propagation shows advantages. The\nmis-classification error of non-linear propagation can be significantly smaller than that of linear\npropagation with sufficiently informative node attributes.\n* When there is a distribution shift of the node attributes between the training and testing datasets,\nnon-linearity provides better transferability in the regime of informative node attributes.\n\n## Run Real Dataset Examples\nWe provide examples with minimal code to run real dataset experiments in `./Neurips2022_Understanding_Non_linearity_in_Graph_Neural_Networks_from_the_Bayesian_Inference_Perspective_Experiments.ipynb`, which includes experiemnts on PubMed, Cora, Citeseer under both Gaussian and Laplacian assumptions.\n\n**Colab:** \u003ca href=\"https://colab.research.google.com/drive/1dkcA2HheSy7y5ivtv2Esc_pjbxwOOin5?usp=sharing\"\u003e\u003cimg src=\"https://colab.research.google.com/assets/colab-badge.svg\" alt=\"Colab\"\u003e\u003c/a\u003e to play with `Neurips2022_Understanding_Non_linearity_in_Graph_Neural_Networks_from_the_Bayesian_Inference_Perspective_Experiments.ipynb` in Colab.\n\n\n## Reference\n\nIf you find our paper and repo useful, please cite our paper:\n```bibtex\n@article{wei2022understanding,\n  title={Understanding Non-linearity in Graph Neural Networks from the Bayesian-Inference Perspective},\n  author={Wei, Rongzhe and Yin, Haoteng and Jia, Junteng and Benson, Austin R and Li, Pan},\n  journal={Advances in Neural Information Processing Systems},\n  year={2022}\n}\n```\n","project_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fgraph-com%2Fbayesian_inference_based_gnn","html_url":"https://awesome.ecosyste.ms/projects/github.com%2Fgraph-com%2Fbayesian_inference_based_gnn","lists_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fgraph-com%2Fbayesian_inference_based_gnn/lists"}