基于自编码器的贝叶斯网嵌入及概率推理
作者:
作者简介:

杜斯(1997-),女,硕士,主要研究领域为大数据分析,不确定性人工智能;祁志卫(1987-),男,博士,CCF学生会员,主要研究领域为大数据分析,不确定性人工智能;岳昆(1979-),男,博士,教授,博士生导师,CCF高级会员,主要研究领域为大数据分析,不确定性人工智能;段亮(1986-),男,博士,副教授,CCF专业会员,主要研究领域为机器学习,大数据分析;王笳辉(1996-),男,博士生,主要研究领域为大数据分析,不确定性人工智能.

通讯作者:

岳昆,E-mail:kyue@ynu.edu.cn

中图分类号:

TP18

基金项目:

国家自然科学基金(62002311);云南省基础研究计划杰出青年项目(2019FJ011);云南省重大科技专项(202002AD080002);云南省基础研究项目(202001BB050052)


Autoencoder-based Bayesian Network Embedding and Probabilistic Inferences
Author:
  • 摘要
  • | |
  • 访问统计
  • |
  • 参考文献 [28]
  • |
  • 相似文献 [20]
  • | | |
  • 文章评论
    摘要:

    贝叶斯网(BN)是不确定性知识表示和推理的基本框架, 广泛用于社交网络、知识图谱和医疗诊断等领域. 特定领域中基于BN的分析诊断和决策支持, 其核心计算任务是基于BN进行多次概率推理. 然而, 使用传统的概率推理方法, 基于同一BN的多次概率推理其中间过程存在很多重复的计算结果, 具有较高的时间复杂度. 为了提高多次概率推理的效率, 提出易于重用和易于计算的贝叶斯网嵌入及相应的概率推理方法. 首先, 借鉴图嵌入的基本思想, 使用点互信息矩阵来表示BN的有向无环图结构和条件概率参数, 提出基于自编码器和注意力机制的BN嵌入方法. 其中, 自编码器的每一编码层利用节点与其邻居节点(父节点和子节点)的相关性生成节点嵌入, 从而在嵌入向量中保存BN节点间的概率依赖关系. 然后, 使用嵌入向量之间的距离来度量节点之间的联合概率, 提出基于嵌入向量的BN概率推理方法. 实验证明, 针对BN的多次概率推理, 所提方法的效率高于现有方法, 且能得到准确的推理结果.

    Abstract:

    Bayesian network (BN), as a preliminary framework for representing and inferring uncertain knowledge, is widely used in social network, knowledge graph, medical diagnosis, etc. The centric computing task of BN-based analysis, diagnosis, and decision-support in specific fields includes multiple probabilistic inferences. However, the high time complexity is doomed on the same BN by using the traditional inference methods, due to the several intermediate results of probability calculations that cannot be shared and reused among different inferences. Therefore, to improve the overall efficiency of multiple inferences on the same BN, this study proposes the method of BN embedding and corresponding probabilistic inferences. First, by incorporating the idea of graph embedding, the study proposes a BN embedding method based on the autoencoder and attention mechanism by transforming BN into the point mutual information matrix to preserve the directed a cyclic graph and conditional probability parameters simultaneously. Specifically, each coding layer of the autoencoder generates node embedding by using the correlation between a node and its neighbors (parent and child nodes) to preserve the probabilistic dependencies. Then, the method for probabilistic inferences to measure the joint probability by using the distance between embedding vectors is proposed. Experimental results show that the proposed method outperforms other state-of-the-art methods in efficiency, achieving accurate results of probabilistic inferences.

    参考文献
    [1] Shen Y, Zhang LZ, Zhang J, Yang M, Tang BZ, Li YL, Lei K. CBN: Constructing a clinical Bayesian network based on data from the electronic medical record. Journal of Biomedical Informatics, 2018, 88: 1-10. [doi: 10.1016/j.jbi.2018.10.007]
    [2] Yue K, Wu H, Fu XD, Xu J, Yin ZD, Liu WY. A data-intensive approach for discovering user similarities in social behavioral interactions based on the Bayesian network. Neurocomputing, 2017, 219: 364-375. [doi: 10.1016/j.neucom.2016.09.042]
    [3] Sun JN, Guo W, Zhang DC, Zhang YX, Regol F, Hu YC, Guo HF, Tang RM, Yuan H, He XQ, Coates M. A framework for recommending accurate and diverse items using Bayesian graph convolutional neural networks. In: Proc. of the 2020 ACM SIGKDD Int’l Conf. on Knowledge Discovery & Data Mining. ACM, 2020. 2030-2039.
    [4] 官赛萍, 靳小龙, 贾岩涛, 王元卓, 程学旗. 面向知识图谱的知识推理研究进展. 软件学报, 2018, 29(10): 2966-2994. http://www.jos.org.cn/1000-9825/5551.htm
    Guan SP, Jin XL, Jia YT, Wang YZ, Cheng XQ. Knowledge reasoning over knowledge graph: A survey. Ruan Jian Xue Bao/Journal of Software, 2018, 29(10): 2966-2994 (in Chinese with English abstract). http://www.jos.org.cn/1000-9825/5551.htm
    [5] Cooper GF. The computational complexity of probabilistic inference using Bayesian belief networks. Artificial Intelligence, 1990, 42(2-3): 393-405. [doi: 10.1016/0004-3702(90)90060-D]
    [6] Dagum P, Luby M. Approximating probabilistic inference in Bayesian belief networks is NP-hard. Artificial Intelligence, 1993, 60(1): 141-153. [doi: 10.1016/0004-3702(93)90036-B]
    [7] . 祁志卫, 王笳辉, 岳昆, 乔少杰, 李劲. 图嵌入方法与应用: 研究综述. 电子学报, 2020, 48(4): 808-818. [doi: 10.3969/j.issn.0372-2112.2020.04.023]
    Qi ZW, Wang JH, Yue K, Qiao SJ, Li J. Methods and applications of graph embedding: A survey. Acta Electronica Sinica, 2020, 48(4): 808-818 (in Chinese with English abstract). [doi: 10.3969/j.issn.0372-2112.2020.04.023]
    [8] Goyal P, Ferrara E. Graph embedding techniques, applications, and performance: A survey. Knowledge-Based Systems, 2018, 151: 78-94. [doi: 10.1016/j.knosys.2018.03.022]
    [9] Shi C, Hu BB, Zhao WX, Yu PS. Heterogeneous information network embedding for recommendation. IEEE Transactions on Knowledge and Data Engineering, 2019, 31(2): 357-370. [doi: 10.1109/TKDE.2018.2833443]
    [10] Tu CC, Liu H, Liu ZY, Sun MS. CANE: Context-aware network embedding for relation modeling. In: Proc. of the 55th Annual Meeting of the Association for Computational Linguistics. Vancouver: ACL, 2017. 1722-1731.
    [11] Liao LZ, He XN, Zhang HW, Chua TS. Attributed social network embedding. IEEE Transactions on Knowledge and Data Engineering, 2018, 30(12): 2257-2270. [doi: 10.1109/TKDE.2018.2819980]
    [12] Shen X, Chung FL. Deep network embedding for graph representation learning in signed networks. IEEE Transactions on Cybernetics, 2020, 50(4): 1556-1568. [doi: 10.1109/TCYB.2018.2871503]
    [13] Wang DX, Cui P, Zhu WW. Structural deep network embedding. In: Proc. of the 2016 ACM SIGKDD Int’l Conf. on Knowledge Discovery and Data Mining. San Francisco: ACM, 2016. 1225-1234.
    [14] Qi ZW, Yue K, Duan L, Wang JH, Qiao SJ, Fu XD. Matrix factorization based Bayesian network embedding for efficient probabilistic inferences. Expert Systems with Applications, 2021, 169: 114294. [doi: 10.1016/j.eswa.2020.114294]
    [15] Levy O, Goldberg Y. Neural word embedding as implicit matrix factorization. In: Proc. of the 27th Int’l Conf. on Neural Information Processing Systems. Montreal: ACM, 2014. 2177-2185.
    [16] Lee JB, Rossi RA, Kim S, Ahmed NK, Koh E. Attention models in graphs: A survey. ACM Transactions on knowledge Discovery from Data, 2019, 13(6): 62. [doi: 10.1145/3363574]
    [17] Perozzi B, Al-Rfou R, Skiena S. DeepWalk: Online learning of social representation. In: Proc. of the 20th ACM SIGKDD Int’l Conf. on Knowledge Discovery and Data Mining. New York: ACM, 2014. 701-710.
    [18] Grover A, Leskovec J. Node2vec: Scalable feature learning for networks. In: Proc. of the 22nd ACM SIGKDD Int’l Conf. on Knowledge Discovery and Data Mining. San Francisco: ACM, 2016. 855-864.
    [19] Mahdavi S, Khoshraftar S, An AJ. dynnode2vec: Scalable dynamic network embedding. In: Proc. of the 2018 IEEE Int’l Conf. on Big Data. Seattle: IEEE, 2018. 3762-3765.
    [20] Singer U, Guy I, Radinsky K. Node embedding over temporal graphs. In: Proc. of the 28th Int’l Joint Conf. on Artificial Intelligence. Macao: IJCAI.org, 2019. 1-8.
    [21] Cao SS, Lu W, Xu QK. Deep neural networks for learning graph representations. In: Proc. of the 30th AAAI Conf. on Artificial Intelligence. Phoenix: ACM, 2016. 1145-1152.
    [22] Park J, Lee M, Chang HJ, Lee K, Choi JY. Symmetric graph convolutional autoencoder for unsupervised graph representation learning. In: Proc. of the 2019 IEEE/CVF Int’l Conf. on Computer Vision. Seoul: IEEE, 2019. 6519-6528.
    [23] Sankar A, Wu YH, Gou L, Zhang W, Yang H. DySAT: Deep neural representation learning on dynamic graphs via self-attention networks. In: Proc. of the 13th Int’l Conf. on Web Search and Data Mining. Houston: ACM, 2020. 519-527.
    [24] Hu FY, Zhu YQ, Wu S, Huang WR, Wang L, Tan TN. GraphAIR: Graph representation learning with neighborhood aggregation and interaction. Pattern Recognition, 2021, 112: 107745. [doi: 10.1016/j.patcog.2020.107745]
    [25] Butz CJ, Oliveira JS, Madsen AL. Bayesian network inference using marginal trees. Int’l Journal of Approximate Reasoning, 2016, 68: 127-152. [doi: 10.1016/j.ijar.2015.07.006]
    [26] Terenin A, Simpson D, Draper D. Asynchronous Gibbs sampling. In: Proc. of the 23rd Int’l Conf. on Artificial Intelligence and Statistics. Palermo: PMLR, 2020. 144-154.
    引证文献
    网友评论
    网友评论
    分享到微博
    发 布
引用本文

杜斯,祁志卫,岳昆,段亮,王笳辉.基于自编码器的贝叶斯网嵌入及概率推理.软件学报,2023,34(10):4804-4820

复制
分享
文章指标
  • 点击次数:632
  • 下载次数: 2169
  • HTML阅读次数: 1264
  • 引用次数: 0
历史
  • 收稿日期:2021-10-01
  • 最后修改日期:2021-12-22
  • 在线发布日期: 2023-04-04
  • 出版日期: 2023-10-06
文章二维码
您是第20040774位访问者
版权所有:中国科学院软件研究所 京ICP备05046678号-3
地址:北京市海淀区中关村南四街4号,邮政编码:100190
电话:010-62562563 传真:010-62562533 Email:jos@iscas.ac.cn
技术支持:北京勤云科技发展有限公司

京公网安备 11040202500063号