[1]楼振凯,侯福均,楼旭明.基于极大似然的马尔可夫链初始状态估计的数学规划[J].南京师范大学学报(自然科学版),2019,42(02):50-56.[doi:10.3969/j.issn.1001-4616.2019.02.008]
　Lou Zhenkai,Hou Fujun,Lou Xuming.Mathematical Programming Models for Estimating the InitialStates of Markov Chains by Maximum Likelihood[J].Journal of Nanjing Normal University(Natural Science Edition),2019,42(02):50-56.[doi:10.3969/j.issn.1001-4616.2019.02.008]
点击复制

基于极大似然的马尔可夫链初始状态估计的数学规划()
分享到：

《南京师范大学学报》（自然科学版）[ISSN:1001-4616/CN:32-1239/N]

卷:
第42卷
期数:
2019年02期
页码:
50-56
栏目:
·数学与计算机科学·
出版日期:
2019-06-30

文章信息/Info

Title:
Mathematical Programming Models for Estimating the InitialStates of Markov Chains by Maximum Likelihood
文章编号:
1001-4616(2019)02-0050-07
作者:
楼振凯1; 侯福均1; 楼旭明2
(1.北京理工大学管理与经济学院,北京 100081)(2.西安邮电大学经济与管理学院,陕西 西安 710121)
Author(s):
Lou Zhenkai1; Hou Fujun1; Lou Xuming2
(1. School of Management and Economics,Beijing Institute of Technology,Beijing 100081,China)(2.School of Economics and Management,Xi’an University of Posts and Telecommunications,Xi’an 710121,China)
关键词:
初始状态估计; 极大似然; 数学规划模型; K-T条件
Keywords:
estimation of the initial state; maximum likelihood; mathematical programming models; K-T condition
分类号:
O221
DOI:
10.3969/j.issn.1001-4616.2019.02.008
文献标志码:
A
摘要:
参数估计是马尔可夫模型中的常见问题. 基于初始状态的重要性,本文对初始状态未知的马尔可夫链模型的初始状态进行估计,并根据状态可见与否将模型分成一般马尔可夫模型和隐马尔可夫模型. 考虑观测状态或观测符号的数量,基于极大似然原理分别建立了线性规划和非线性规划模型,并证明各阶段状态的概率满足规范性. 对于线性规划模型,指出其可以用单纯形法求解,并给出了解的表达. 对于非线性模型,指出其最优解的存在性,并利用库恩-塔克条件(K-T条件)将模型转化成方程组的形式. 算例分析中,在基于库恩-塔克条件的方程组不易求解的情形下,运用lingo得到了满足模型的解.
Abstract:
Parameters estimationis a common issue in Markov models. Owing to the importance of the initial state,in this paper we estimate the initial state for Markov chain models which the initial state is unknown. According to whether the states are visible,we divide the models into general Markov models and hidden Markov models. We build linear programming models and non-linear programming models by maximum likelihood with considering the amount of states or observation symbols,and prove that the probabilities of state at each stage meet the normalization of probability. For linear programming models,we point out that they can be solved by the simplex method,and show the expression of solution. For non-linear programming models,we account for the existence of the optimal solution,and turn models into equations by K-T condition. In the examples,we apply lingo to obtain the optimal solution while the equations are hard to solve.

参考文献/References:

[1] ARNAB N,LAURENT E G. Robust control of Markov decision processes with uncertain transition matrices[J]. Operations research,2005,53(5):780-798.
[2]ERICK D,SHIE M. Percentile optimization for Markov decision processes with parameter uncertainty[J]. Operations research,2009,58(1):203-213.
[3]YU S Z,HISASHI K. An efficient forward-backward algorithm for an explicit-duration hidden Markov model[J]. IEEE signal processing letters,2003,10:11-14.
[4]XIE L,UGRINOVSKII V A,PETERSEN I R. Petersen. Finite horizon robust state estimation for uncertain finite alphabet hidden Markov models with conditional relative entropy constraints[J]. SIAM journal on control and optimization,2008,4(1):4497-4502.
[5]JAMIE S E,VIKRAM K. Hidden Markov model state estimation with randomly delayed observations[J]. IEEE transactions on signal processing,1999,47(8):2157-2166.
[6]LIANG J L,WANG Z D,LIU X H. Distributed state estimation for uncertain Markov-type sensor networks with mode-dependent distributed delays[J]. International journal of robust and nonlinear control,2012,22(3):331-346.
[7]SHI D W,ROBERT J E,CHEN T W. Event-based state estimation of discrete-state hidden Markov models[J]. Automatica,2016,65:12-26.
[8]BRIAN G L. Maximum-likelihood estimation for hidden Markov models[J]. Stochastic processes and their applications,1992,40:127-143.
[9]SUN Q,LIM C C,LIU F. Maximum likelihood state estimation for Markov jump systems with uncertain mode-dependent delays[J]. Journal of the Franklin institute,2016,353:594-614.
[10]张虎,胡淑兰. 马尔可夫转换模型的极大似然估计的算法[J]. 统计与决策,2011,6:26-27.
[11]RICHARD Z,ZHOU M C. Petri nets and industrial applications:a tutorial[J]. IEEE transactions on industrial electronics,1994,41(6):567-583.
[12]PHUC D V,ANNE B,CHRISTOPHE B. Reliability importance analysis of Markovian systems at stead state using perturbation analysis[J]. Reliability engineering and system safety,2013,93(11):1605-1615.
[13]荣腾中,肖智. 高阶马尔可夫链平稳分布的存在唯一性[J]. 系统工程理论与实践,2013,33(8):2015-2020.
[14]MEDER Z Z,FLESCH J,PEETERS R. Optimal choice for finite and infinite horizons[J]. Operations research letters,2012,40(6):469-474.
[15]MACDONALD I,ZUCCHINI W. Hidden Markov and other models for discrete-valued time series[M]. London:Chapman and Hall,1997.
[16]CHING W K,MICHAEL K N. Markov chains:models,algorithms and applications[M]. New York:Publications of the American Statistical Association,2006.
[17]刘克,曹平. 马尔可夫决策过程理论与应用[M]. 北京:科学出版社,2015.

备注/Memo

备注/Memo:
收稿日期:2019-01-07.
基金项目:国家自然科学基金面上项目(71571019).
通讯联系人:侯福均,副教授,博士生导师,研究方向:群决策、博弈论. E-mail:houfj@bit.edu.cn
更新日期/Last Update: 2019-06-30