多维贝叶斯网络分类器结构学习算法

doi:10.11772/j.issn.1001-9081.2014.04.1083

计算机应用 ›› 2014, Vol. 34 ›› Issue (4): 1083-1088.DOI: 10.11772/j.issn.1001-9081.2014.04.1083

多维贝叶斯网络分类器结构学习算法

傅顺开,Sein Minn,李志强

华侨大学计算机科学与技术学院,福建厦门 361021

收稿日期:2013-09-30 修回日期:2013-12-21 出版日期:2014-04-01 发布日期:2014-04-29
通讯作者: 傅顺开
作者简介:傅顺开(1978-),男,福建仙游人,讲师,博士,主要研究方向:数据挖掘、移动互联网;
Sein Minn(1990-),男,缅甸人,硕士研究生,主要研究方向:数据挖掘;
李志强(1990-),男,福建安溪人,硕士研究生,主要研究方向:数据挖掘。
基金资助:
国家自然科学基金资助项目;中央高校基本科研业务费专项资金资助项目;厦门科技计划基金资助项目;华侨大学科研基金资助项目

Structure learning algorithm for general multi-dimensional Bayesian network classifiers

FU Shunkai,SEIN Minn,LI Zhiqiang

College of Computer Science and Technology, Huaqiao University, Xiamen Fujian 361021, China

Received:2013-09-30 Revised:2013-12-21 Online:2014-04-01 Published:2014-04-29
Contact: FU Shunkai
Supported by:
; the Scientific Research Foundation of Huaqiao University

摘要/Abstract

摘要：

传统多维贝叶斯网络分类器(MBNC)限制其模型结构必须是二分的,通过移除该限制可得到更准确的对关联分布建模的通用MBNC(GMBNC)。基于局部马尔可夫毯的迭代搜索，提出可准确学习GMBNC的算法IPC-GMBNC。该算法由于无需学习全局贝叶斯网络(BN),可扩展性强。基于已知贝叶斯网络模型而随机生成的数据上所执行的实验显示,IPC-GMBNC可有效推导出目标结构；而且与传统的全局结构学习算法PC相比,IPC-GMBNC可节省大量的计算量。

Abstract:

The conventional Multi-dimensional Bayesian Network Classifier (MBNC) requires its structure be bi-partitie. Removing this constraint can result into a new tool named General MBNC (GMBNC), and it enables us to model the underlying joint distribution more correctly. Based on iterative local search of Markov blankets, an algorithm called IPC-GMBNC was proposed to induce the exact structure of GMBNC. The proposed algorithm has good scalability because it does not need to recover the global Bayesian Network (BN) first. The experiments on samples generated from known Bayesian network structures indicate that IPC-GMBNC is effective, and it brings great reduction on computing complexity compared to global search approach, e.g. PC algorithm.

中图分类号:

TP391

傅顺开 Sein Minn 李志强. 多维贝叶斯网络分类器结构学习算法[J]. 计算机应用, 2014, 34(4): 1083-1088.

FU Shunkai SEIN Minn LI Zhiqiang. Structure learning algorithm for general multi-dimensional Bayesian network classifiers[J]. Journal of Computer Applications, 2014, 34(4): 1083-1088.

参考文献

［1］CHICKERING D M, GEIGER D, HECKERMAN D. LearningBayesian network is NP-hard, MSR-TR-94-17 ［R］. [S.l.]: Microsoft Research, 1994.
［2］CHOW C K, LIU C N. Approximating discrete probability distributions with dependence trees ［J］. IEEE Transactions on Information Theory, 1968, 14(3): 462-467.
［3］de WAAL P R, van der GAAG L C. Inference and learning inmulti-dimensional Bayesian network classifiers ［C］// Proceedings of the 9th European Conferences on Symbolic and Quantitative Approaches to Reasoning with Uncertainty, LNCS 4724. Berlin: Springer-Verlag, 2007: 501-511.
［4］REBANE G, PEARL J. The recovery of causal poly-trees from statistical data ［C］// UAI '87: Proceedings of the 3rd Conference on Uncertainty in Artificial Intelligence. New York: Elsevier Science Inc., 1987: 175-182.
［5］ZARAGOZA J C, SUCAR E, MORALES E. A two-step method to learn multidimensional Bayesian network classifiers based on mutual information measures ［C］// Proceedings of the 24th International FLAIRS Conference. Menlo Park, California: AAAI Press, 2011:644-649.
［6］RODRIGUEZ J D, LOZANO J A. Multi-objective learning of multi-dimensional Bayesian classifiers ［C］// HIS '08: Proceedings of the 8th International Conference on Hybrid Intelligent Systems. Washington, DC: IEEE Computer Society, 2008: 501-506.
［7］BIELZ C, LI G, LARRANGA P. Multi-dimensional classification with Bayesian networks ［J］. International Journal of Approximate Reasoning, 2011, 52(6): 705-727.
［8］BORCHANI H, BIELZA C, LARRANGA P. Learning CB-decomposable multi-dimensional Bayesian network classifiers ［C］// PGM'10: Proceedings of the 5th European Workshop on Probabilistic Graphical Models. Helsinki: HIIT Publications, 2010: 23-32.
［9］ZARAGOZA J H, SUCAR L E, MORALES E F, et al.Bayesian chain classifiers for multidimensional classification ［C］// IJCAI '11: Proceedings of the 22nd International Joint Conference on Artificial Intelligence. Menlo Park, California: AAAI Press, 2011, 3: 2192-2197.
［10］BORCHANI H, BIELZA C, MARTNEZ-ARTN P, et al.Markov blanket-based approach for learning multi-dimensional Bayesian network classifiers: an application to predict the European Quality of life-5 Dimensions (EQ-5D) from the 39-item Parkinson's Disease Questionnaire (PDQ-39) ［J］. Journal of Biomedical Informatics, 2012, 45(6): 1175-1184.
［11］BORCHANI H, BIELZA C, TORO C, et al.Predicting human immunod eficiency virus inhibitors using multi-dimensional Bayesian network classifiers ［J］. Artificial Intelligence in Medicine, 2013, 57(3): 219-229.
［12］ALIFERIS C F, TSAMARDINOS I, STATNIKOV A. HITON: a novel Markov blanket algorithm for optimal variable selection ［C］// Proceedings of the 2003 Annual Symposium on American Medical Informatics Association (AMIA). Washington, DC: AMIA Publications, 2003: 21-25.
［13］PENA J M, NILSSON R, BJORKEGREN J, et al.Towards scalable and data efficient learning of Markov boundaries ［J］. International Journal of Approximate Reasoning, 2007, 45(2):211-232.
［14］VERMA T, PEAL J. Equivalence and synthesis of causal models ［C］// UAI '90: Proceedings of the 6th Annual Conference on Uncertainty in Artificial Intelligence. New York: Elsevier Science Inc., 1990: 255-268.
［15］van der GAAG L C, de WAAL P R. Multi-dimensional Bayesian network classifiers, UU-CS-2006-056 ［R］. Utrecht: Utrecht University, Department of Information and Computing Sciences, 2006.
［16］FU S, DESMARAIS M C. Tradeoff analysis of different Markov blanket local learning approaches ［C］// PAKDD '08: Proceedings of the 12th Pacific-Asia Conference on Advances in Knowledge Discovery and Data Mining. Berlin: Springer-Verlag, 2008: 562-571.
［17］FU S, DESMARAIS M C. Fast Markov blanket discovery algorithm via local learning within single pass ［C］// Canadian AI '08: Proceedings of the 21st Conference of the Canadian Society for Computational Studies of Intelligence. Berlin: Springer-Verlag, 2008: 96-107.
［18］FU S, DESMARAIS M C. Markov blanket based feature selection: a review of past decade ［C］// Proceedings of the 2010 World Congress on Engineering, LNCS 2183. Berlin: Springer-Verlag, 2010: 321-328.
［19］PEARL J. Causality: models, reasoning, and inference ［M］.Cambridge: Cambridge University Press, 2000.
［20］SPIRTES P, GLYMOUR C, SCHEINES R. Causation, prediction and search ［M］. 2nd ed. Cambridge: MIT Press, 2001.

[1]	章悦, 张亮, 谢非, 杨嘉乐, 张瑞, 刘益剑. 基于实例分割模型优化的道路抛洒物检测算法[J]. 《计算机应用》唯一官方网站, 2021, 41(11): 3228-3233.
[2]	李凯, 李洁. 基于pinball损失的结构模糊多分类支持向量机算法[J]. 《计算机应用》唯一官方网站, 2021, 41(11): 3104-3112.
[3]	胡誉生, 何炳蔚, 邓清康. 混合视觉系统的运动物体检测和静态地图重建[J]. 《计算机应用》唯一官方网站, 2021, 41(11): 3332-3336.
[4]	高洁, 朱元, 陆科. 基于雷达和相机融合的目标检测方法[J]. 《计算机应用》唯一官方网站, 2021, 41(11): 3242-3250.
[5]	彭博, 罗娅茹, 谢盛华, 尹立雪. 联合深度学习的通用血流向量成像方法[J]. 《计算机应用》唯一官方网站, 2021, 41(11): 3368-3375.
[6]	陈吉成, 陈鸿昶. 基于张量建模和进化K均值聚类的社区检测方法[J]. 《计算机应用》唯一官方网站, 2021, 41(11): 3120-3126.
[7]	张嘉琪, 张月琴, 陈健. 优化强化学习路径特征分类的脉象识别法[J]. 《计算机应用》唯一官方网站, 2021, 41(11): 3402-3408.
[8]	任俊伟, 曾诚, 肖丝雨, 乔金霞, 何鹏. 基于会话的多粒度图神经网络推荐模型[J]. 《计算机应用》唯一官方网站, 2021, 41(11): 3164-3170.
[9]	孙琳, 袁玉波. 基于人眼状态的瞌睡识别算法[J]. 《计算机应用》唯一官方网站, 2021, 41(11): 3213-3218.
[10]	葛晨宇, 董良, 许伊昆, 常毅, 张宏鸣. 基于总变分低秩组稀疏的全球雷达数据修复算法[J]. 《计算机应用》唯一官方网站, 2021, 41(11): 3353-3361.
[11]	闫钧华, 侯平, 张寅, 吕向阳, 马越, 王高飞. 基于多尺度多分类器卷积神经网络的混合失真类型判定方法[J]. 《计算机应用》唯一官方网站, 2021, 41(11): 3178-3184.
[12]	李福海, 蒋慕蓉, 杨磊, 谌俊毅. 基于生成对抗网络的梯度引导太阳斑点图像去模糊方法[J]. 《计算机应用》唯一官方网站, 2021, 41(11): 3345-3352.
[13]	曹建芳, 闫敏敏, 贾一鸣, 田晓东. 融合迁移学习的Inception-v3模型在古壁画朝代识别中的应用[J]. 《计算机应用》唯一官方网站, 2021, 41(11): 3219-3227.
[14]	刘太亨, 何昭水. 基于自编码和知识蒸馏的表面缺陷检测方法[J]. 《计算机应用》唯一官方网站, 2021, 41(11): 3200-3205.
[15]	张阳, 王小宁. 基于Word2Vec词嵌入和高维生物基因选择遗传算法的文本特征选择方法[J]. 《计算机应用》唯一官方网站, 2021, 41(11): 3151-3155.

多维贝叶斯网络分类器结构学习算法

Structure learning algorithm for general multi-dimensional Bayesian network classifiers

PDF

可视化

摘要/Abstract

引用本文

使用本文

参考文献

相关文章 15

编辑推荐

Metrics