Please wait a minute...
Chin. Phys. B, 2024, Vol. 33(5): 058401    DOI: 10.1088/1674-1056/ad20db
INTERDISCIPLINARY PHYSICS AND RELATED AREAS OF SCIENCE AND TECHNOLOGY Prev   Next  

A novel complex-high-order graph convolutional network paradigm: ChyGCN

He-Xiang Zheng(郑和翔)1, Shu-Yu Miao(苗书宇)2, and Chang-Gui Gu(顾长贵)1,†
1 Business School, University of Shanghai for Science and Technology, Shanghai 200093, China;
2 Ant Group, Shanghai 200000, China
Abstract  In recent years, there has been a growing interest in graph convolutional networks (GCN). However, existing GCN and variants are predominantly based on simple graph or hypergraph structures, which restricts their ability to handle complex data correlations in practical applications. These limitations stem from the difficulty in establishing multiple hierarchies and acquiring adaptive weights for each of them. To address this issue, this paper introduces the latest concept of complex hypergraphs and constructs a versatile high-order multi-level data correlation model. This model is realized by establishing a three-tier structure of complexes-hypergraphs-vertices. Specifically, we start by establishing hyperedge clusters on a foundational network, utilizing a second-order hypergraph structure to depict potential correlations. For this second-order structure, truncation methods are used to assess and generate a three-layer composite structure. During the construction of the composite structure, an adaptive learning strategy is implemented to merge correlations across different levels. We evaluate this model on several popular datasets and compare it with recent state-of-the-art methods. The comprehensive assessment results demonstrate that the proposed model surpasses the existing methods, particularly in modeling implicit data correlations (the classification accuracy of nodes on five public datasets Cora, Citeseer, Pubmed, Github Web ML, and Facebook are 86.1$\pm $0.33, 79.2$\pm $0.35, 83.1$\pm $0.46, 83.8$\pm $0.23, and 80.1$\pm $0.37, respectively). This indicates that our approach possesses advantages in handling datasets with implicit multi-level structures.
Keywords:  raph convolutional network      complex modeling      complex hypergraph  
Received:  20 November 2023      Revised:  10 January 2024      Accepted manuscript online:  22 January 2024
PACS:  84.35.+i (Neural networks)  
  64.60.aq (Networks)  
  47.50.Cd (Modeling)  
  07.05.Mh (Neural networks, fuzzy logic, artificial intelligence)  
Fund: Project supported by the National Natural Science Foundation of China (Grant Nos. 12275179 and 11875042) and the Natural Science Foundation of Shanghai Municipality, China (Grant No. 21ZR1443900).
Corresponding Authors:  Chang-Gui Gu     E-mail:  gu_changgui@163.com

Cite this article: 

He-Xiang Zheng(郑和翔), Shu-Yu Miao(苗书宇), and Chang-Gui Gu(顾长贵) A novel complex-high-order graph convolutional network paradigm: ChyGCN 2024 Chin. Phys. B 33 058401

[1] Shlomi J, Battaglia P and Vlimant J R 2020 Mach. Learn.: Sci. Technol. 2 021001
[2] Kipf T N and Welling M 2016 arxiv: 1609.02907
[3] Defferrard M, Bresson X and Vandergheynst P 2016 NIPS 29
[4] Jin W, Derr T, Wang Y Q, Ma Y and Liu Z T 2021 Proceedings of The 14th ACM International Conference on Web Search and Data Mining pp. 148-156
[5] Bhatti U A, Tang H, Wu G L, Marjan S and Hussain A 2023 Int. J. Intell. Syst. 2023 1
[6] Wu Z H, Pan S R, Chen F W, Long G D, Zhang C Q and Yu P 2020 IEEE. T. Neur. Net. Lear. 32 4
[7] Zhang L and Lu H P 2020 Proceedings of The 29th ACM International Conference on Information & Knowledge Management 2020 pp. 1813- 1822
[8] Hamilton W, Ying Z and Leskovec J 2017 NIPS 30
[9] Gasteiger J, Bojchevski A and Günnemann S 2018 arxiv: 1810.05997
[10] Wang X, Ji H Y, Shi C, Wang B, Ye Y F, Cui P and Yu P 2019 The World Wide Web Conference 2022-2032
[11] Xu K, Hu W H, Leskovec J and Jegelka S 2018 arXiv: 1810.00826
[12] Verma V, Qu M, Kawaguchi K, Lamb A, Bengio Y, Kannala J and Tang J 2021 Proceedings of the AAAI Conference on Artificial Intelligence 35 10024
[13] Zeng H Q, Zhou H K, Srivastava A, Kannan R and Prasanna V 2019 arXiv: 1907.04931
[14] Deng C H, Zhao Z Q, Wang Y Y, Zhang Z R and Feng Z 2020 arXiv: 1910.02370
[15] Abu-El-Haija S, Perozzi B, Kapoor A, Alipourfard N, Lerman K, Harutyunyan H, Steeg G and Galstyan A 2019 International Conference on Machine Learning 97 21
[16] Gao Y, Feng Y F, Ji S Y and Ji R R 2022 IEEE. T. Pattern. Anal. 45 3181
[17] Pham P, Nguyen L, Nguyen N, Kozma R and Vo B 2023 Inform. Sci. 620 105
[18] Xie L Y, Lu Y, Furuhata T, Yamakawa S, Zhang W, Regmi A, Kara L and Shimada K 2022 Comput. Ind. 142 103697
[19] Wang Y, Wang H J, Jin H, Huang X R and Wang X 2021 Inform. Sci. 581 932
[20] Yu L, Sun L L, Du B W, Liu C R, Xiong H and Lv W F 2020 Proceedings of the 26th ACM SIGKDD International Conference on Knowledge Discovery & Data Mining 2020 pp. 1083-1091
[21] Chiang W L, Liu X Q, Si S, Li Y, Bengio S and Hsieh C 2019 Proceedings of The 25th ACM SIGKDD International Conference on Knowledge Discovery & Data Mining pp. 257-266
[22] Ye J C, Liu Z H, Du B W, Sun L L, Li W M, Fu Y J and Xiong H 2022 Proceedings of The 28th ACM SIGKDD Conference on Knowledge Discovery and Data Mining pp. 2296-2306
[23] Vazquez A 2023 Phys. Rev. E 107 024316
[1] RNAGCN: RNA tertiary structure assessment with a graph convolutional network
Chengwei Deng(邓成伟), Yunxin Tang(唐蕴芯), Jian Zhang(张建), Wenfei Li(李文飞), Jun Wang(王骏), and Wei Wang(王炜). Chin. Phys. B, 2022, 31(11): 118702.
No Suggested Reading articles found!