零一膨胀二项混合回归模型的改进EM算法

doi:10.3969/j.issn.1006-2475.2018.03.012

计算机与现代化

零一膨胀二项混合回归模型的改进EM算法

(1.西安航空学院理学院,陕西西安710077;2.西安交通大学经济与金融学院,陕西西安710049)

收稿日期:2017-09-12 出版日期:2018-04-03 发布日期:2018-04-03
作者简介:吕敏红(1985-),女,陕西合阳人,西安航空学院理学院讲师,硕士,研究方向:概率论与数理统计; 闫奕荣(1974-),女,陕西合阳人,西安交通大学经济与金融学院副教授,博士,研究方向:区域经济,国际贸易; 吴成晶(1987-),女,陕西安康人,讲师,硕士,研究方向:初等数论。
基金资助:
国家自然科学基金资助项目(11647050); 陕西省教育厅科研计划项目(16JK1394)

EM Algorithm Amendment of Zero-and-one Inflated Binomial Mixture Regression Model

(1. School of Science, Xi’an Aerotechnical University, Xi’an 710077, China;
2. School of Economics and Finance of Xi’an Jiaotong University, Xi’an 710049, China)

Received:2017-09-12 Online:2018-04-03 Published:2018-04-03

摘要/Abstract

摘要： 首先对存在过多0和1的观测数据提出了零一膨胀混合回归模型，由于EM算法一般会使得估计收敛到局部最优解上，所以提出了一种修正EM算法,对具有有限混合成分的零一膨胀二项回归模型（ZOIB）的参数进行估计。最后通过模拟研究说明该方法的有效性。

关键词: 零一膨胀, 混合回归, 加速MCEM算法

Abstract: This paper proposes a zero-and-one inflated regression model with “heterogeneity”. Since standard EM algorithm will make the parameter estimates converge to local optimum, this paper introduces the MCEM algorithm to address this shortcoming, then the estimation of parameter is setted up for the zero-and-one inflated binomial mixture regression model. Finally, the corresponding simulation is given to illustrate the proposed method.

Key words: zero-and-one inflation, mixture regression, acceleration of MCEM algorithm

中图分类号:

TP309.2

吕敏红1,闫奕荣2,吴成晶1. 零一膨胀二项混合回归模型的改进EM算法[J]. 计算机与现代化, doi: 10.3969/j.issn.1006-2475.2018.03.012.

LYU Min-hong1, YAN Yi-rong2, WU Cheng-jing1. EM Algorithm Amendment of Zero-and-one Inflated Binomial Mixture Regression Model[J]. Computer and Modernization, doi: 10.3969/j.issn.1006-2475.2018.03.012.

参考文献

［1］ Tchetgen E J, Coull B A. A diagnostic test for the mixing distribution in a generalised linear mixed model［J］. Biometrika, 2006,93(4):1003-1010.
［2］ Li Pengfei, Chen Jiahua. Testing the oder of a finite mixture［J］. Journal of American Statistical Association, 2010,105(491):1084-1092.
［3］ Zhu Hongtu, Zhang Heping. Hypothesis testing in mixture regression models［J］. Journal of Royal Statistical Society Series B, 2004,66(1):3-16.
［4］ Lin T I, Lee J C, Yen S Y. Finite mixture modeling using the skew normal distribution［J］. Statistica Sinica, 2007,17(3):909-927.
［5］ Xiang Liming, Yau K K W, Lee A H, et al. Influence diagnostics for two-component Poisson mixture regression models: Applications in public health［J］. Statistics in Medicine, 2005(19):3053-3071.
［6］陈家骅,李鹏飞,谭鲜明. 混合Von Mises模型的参数估计［J］. 系统科学与数学, 2007,27(1):59-67.
［7］ Cohen A. Estimation of the Poisson parameter from truncated samples and from censored samples［J］. Journal of American Statistical Association, 1954,49(265):158-168.
［8］ Lambert D. Zero-inflated Poisson regression with an application to defects in manufacturing［J］. Technometrics, 1992,34(1):1-14.
［9］ Fahrmeir L, Echavarria L O. Structured additive regression for overdispersed and zero-inflated count data［J］. Applied Stochastic Models in Business and Industry, 2006,22(4):351-369.
［10］Ridout M, Hinde J, Demetrio C G B. A score test for testing zero-inflated Poisson regression model against zero-inflated negative binomial alternatives［J］. Biometrics, 2001,57(1):219-223.
［11］Jansakul N, Hinde J P. Score tests for zero-inflated Poission models［J］. Computational Statistics and Data Analysis， 2002,40(1):75-96．
［12］Melkersson M, Olsson C. Is Visiting the Dentist A Good Habit? Analyzing Count Data with Excess Zeros and Excess Ones［D］. Sweden: Umer University, 1999.
［13］Zhang Chi, Tian Guoliang, Wang Kai. Properties of the zero-and-one inflated Poisson distribution and statistical inference methods［J］. Statistics and Its Interface, 2016,9(1):11-32.
［14］Biernacki C. Initializing EM using the properties of its trajectories in Gaussian mixtures［J］. Statistics and Computing, 2004,14(3):267-279.
［15］吕敏红,闫奕荣,杨青. 零一膨胀泊松回归模型的EM算法改进［J］. 河南科学, 2017，35(7):1037-1041.
［16］罗季. Monte Carlo EM加速算法［J］. 应用概率统, 2008,24(3):312-318.
［17］卢玉桂，韦新星，赵丽棉. 多层线性模参数估计的MCEM算法［J］. 数学的实践与认识, 2016,46(11):225-230.

零一膨胀二项混合回归模型的改进EM算法

EM Algorithm Amendment of Zero-and-one Inflated Binomial Mixture Regression Model

可视化

摘要/Abstract

引用本文

使用本文

参考文献

相关文章 1

编辑推荐

Metrics

本文评价