当前位置：文档库 › 概率图模型研究进展综述

概率图模型研究进展综述

软件学报ISSN 1000-9825, CODEN RUXUEW E-mail: jos@https://www.wendangku.net/doc/b67652510.html,

Journal of Software,2013,24(11):2476?2497 [doi: 10.3724/SP.J.1001.2013.04486] https://www.wendangku.net/doc/b67652510.html,

概率图模型研究进展综述

张宏毅1,2, 王立威1,2, 陈瑜希1,2

1(机器感知与智能教育部重点实验室(北京大学),北京 100871)

2(北京大学信息科学技术学院智能科学系,北京 100871)

通讯作者: 张宏毅, E-mail: hongyi.zhang.pku@https://www.wendangku.net/doc/b67652510.html,

摘要: 概率图模型作为一类有力的工具,能够简洁地表示复杂的概率分布,有效地(近似)计算边缘分布和条件分

布,方便地学习概率模型中的参数和超参数.因此,它作为一种处理不确定性的形式化方法,被广泛应用于需要进行

自动的概率推理的场合,例如计算机视觉、自然语言处理.回顾了有关概率图模型的表示、推理和学习的基本概念

和主要结果,并详细介绍了这些方法在两种重要的概率模型中的应用.还回顾了在加速经典近似推理算法方面的新

进展.最后讨论了相关方向的研究前景.

关键词: 概率图模型;概率推理;机器学习

中图法分类号: TP181文献标识码: A

中文引用格式: 张宏毅,王立威,陈瑜希.概率图模型研究进展综述.软件学报,2013,24(11):2476?2497.https://www.wendangku.net/doc/b67652510.html,/

1000-9825/4486.htm

英文引用格式: Zhang HY, Wang LW, Chen YX. Research progress of probabilistic graphical models: A survey. Ruan Jian Xue

Bao/Journal of Software, 2013,24(11):2476?2497 (in Chinese).https://www.wendangku.net/doc/b67652510.html,/1000-9825/4486.htm

Research Progress of Probabilistic Graphical Models: A Survey

ZHANG Hong-Yi1,2, WANG Li-Wei1,2, CHEN Yu-Xi1,2

1(Key Laboratory of Machine Perception (Peking University), Ministry of Education, Beijing 100871, China)

2(Department of Machine Intelligence, School of Electronics Engineering and Computer Science, Peking University, Beijing 100871,

China)

Corresponding author: ZHANG Hong-Yi, E-mail: hongyi.zhang.pku@https://www.wendangku.net/doc/b67652510.html,

Abstract: Probabilistic graphical models are powerful tools for compactly representing complex probability distributions, efficiently

computing (approximate) marginal and conditional distributions, and conveniently learning parameters and hyperparameters in

probabilistic models. As a result, they have been widely used in applications that require some sort of automated probabilistic reasoning,

such as computer vision and natural language processing, as a formal approach to deal with uncertainty. This paper surveys the basic

concepts and key results of representation, inference and learning in probabilistic graphical models, and demonstrates their uses in two

important probabilistic models. It also reviews some recent advances in speeding up classic approximate inference algorithms, followed

by a discussion of promising research directions.

Key words: probabilistic graphical model; probabilistic reasoning; machine learning

我们工作和生活中的许多问题都需要通过推理来解决.通过推理,我们综合已有的信息,对我们感兴趣的未

知量做出估计,或者决定采取某种行动.例如,程序员通过观察程序在测试中的输出判断程序是否有错误以及需

要进一步调试的代码位置,医生通过患者的自我报告、患者体征、医学检测结果和流行病爆发的状态判断患者

可能罹患的疾病.一直以来,计算机科学都在努力将推理自动化,例如,编写能够自动对程序进行测试并且诊断

?基金项目: 国家自然科学基金(61222307, 61075003)

收稿时间:2013-07-17; 修改时间: 2013-08-02; 定稿时间: 2013-08-27