Bayesian analysis of the multivariate Poisson distribution

被引：25

作者：

Tsionas, EG ^{[1
]}

机构：

[1] Council Econ Advisors, Minist Natl Econ, Athens 10180, Greece

来源：

COMMUNICATIONS IN STATISTICS-THEORY AND METHODS | 1999年 / 28卷 / 02期

关键词：

Gibbs sampling; data augmentation; multivariate probability mass function;

D O I：

10.1080/03610929908832305

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：

The paper presents a Bayesian analysis of the multivariate Poisson distribution, based on Gibbs sampling with data augmentation. This computational approach to the Bayesian integration problem, avoids explicit computation of the probability mass function via approximations or recursive calculations. The new method is shown to perform well in artificial examples as well as in an environmental pollution data set.

引用

页码：431 / 451

页数：21

共 46 条

[1] A BAYESIAN-ANALYSIS OF A POISSON RANDOM EFFECTS MODEL FOR HOME RUN HITTERS [J].

ALBERT, J .

AMERICAN STATISTICIAN, 1992, 46 (04) :246-253

[2]

Bahadur R, 1961, STUDIES ITEM ANAL PR, V6, P158

[3]

BALAKRISHNAN N, 1998, DISTRIBUTIONS STAT D

[4]

BELOV AG, 1993, VESTNIK MGU 15, P58

[5]

BENNETT JE, 1996, MARKOV CHAIN MONTE C

[6] DYNAMIC COUNT DATA MODELS OF TECHNOLOGICAL INNOVATION [J].

BLUNDELL, R ;

GRIFFITH, R ;

VANREENEN, J .

ECONOMIC JOURNAL, 1995, 105 (429) :333-344

[7]

BREYER LA, 1998, UNPUB METROPOLIS DIF

[8]

CAMPBELL JT, 1938, P EDINBURGH MATH S 2, V4, P18

[9]

CHIB S, 1998, UNPUB J ECONOMETRICS

[10] ON INFINITELY DIVISIBLE RANDOM VECTORS [J].

DWASS, M ;

TEICHER, H .

ANNALS OF MATHEMATICAL STATISTICS, 1957, 28 (02) :461-470

← 1 2 3 4 5 →