The Monte Carlo EM method for estimating multinomial probit latent variable models

被引：1
作者：
Zhou, Xingcai [2 ,3 ]
Liu, Xinsheng [1 ,2 ]
机构：
[1] Nanjing Univ Aeronaut & Astronaut, Inst Nanosci, Nanjing 210016, Peoples R China
[2] Nanjing Univ Aeronaut & Astronaut, Coll Sci, Nanjing 210016, Peoples R China
[3] Coll Tongling, Anhua 244000, Peoples R China
来源：
COMPUTATIONAL STATISTICS | 2008年 / 23卷 / 02期
关键词：
multinomial probit model; latent variable; maximum likelihood estimates; Monte Carlo EM; factor analysis;
D O I：
10.1007/s00180-007-0091-7
中图分类号：
O21 [概率论与数理统计]; C8 [统计学];
学科分类号：
020208 ; 070103 ; 0714 ;
摘要：
We propose a multinomial probit (MNP) model that is defined by a factor analysis model with covariates for analyzing unordered categorical data, and discuss its identification. Some useful MNP models are special cases of the proposed model. To obtain maximum likelihood estimates, we use the EM algorithm with its M-step greatly simplified under Conditional Maximization and its E-step made feasible by Monte Carlo simulation. Standard errors are calculated by inverting a Monte Carlo approximation of the information matrix using Louis's method. The methodology is illustrated with a simulated data.
引用
收藏
页码：277 / 289
页数：13
相关论文
共 33 条
[11]  
GEWEKE J, 1994, 177 FED RES BANK MIN
[12]  
HAJIVASSILIOU V, 1992, 1021 YAL U
[13]  
HARTLEY H, 1958, BIOMETRICS, V14, P179
[14]  
HAUSMAN J, 1978, ECONOMETRICA, V45, P319
[15]  
JORESKOG KG, 1979, ADV FACTOR ANAL ST 1
[16]   A NOTE ON IDENTIFICATION IN THE MULTINOMIAL PROBIT MODEL [J].
KEANE, MP .
JOURNAL OF BUSINESS & ECONOMIC STATISTICS, 1992, 10 (02) :193-200
[17]   Bayesian analysis of structural equation models with dichotomous variables [J].
Lee, SY ;
Song, XY .
STATISTICS IN MEDICINE, 2003, 22 (19) :3073-3088
[18]  
Lerman S., 1981, STRUCTURAL ANAL DISC, V10, P305
[19]  
LIU CH, 1995, STAT SINICA, V5, P19
[20]  
LOUIS TA, 1982, J ROY STAT SOC B MET, V44, P226
1234