An EM type algorithm for maximum likelihood estimation is proposed for the case of
mixed Poisson regression models. The algorithm makes use of the mixture
representation of such models. Two members of this family are examined in depth, the
negative binomial regression model and the Poisson-inverse Gaussian regression
model. Closed form expressions are derived for both models leading to easily
programmable algorithms. Especially for the case of the Poisson-inverse Gaussian
model no special numerical techniques are needed. The algorithms are applied to a
real data set concerning crime data from Greece.
Abramowitz M
, and
Stegum IA
(1974) Handbook of mathematical functions.
New York: Dover
.
2.
Aitkin M
(1996)
A general maximum likelihood analysis of overdispersion in generalized
linear models
. Statistics and Computing, 6,
251-262
.
3.
Aitkin M
(1999)
A general maximum likelihood analysis of variance components in generalized
linear models
. Biometrics, 55,
117-128
.
4.
Anderson DA
, and
Hinde J
(1988)
Random effects in generalized linear models and the EM algorithm
. Communications in Statistics - Theory and Methods,
17,
3847-3856
.
5.
Athreya K
(1986)
Another conjugate family for the normal distribution
. Statistics and Probability Letters, 4,
61-64
.
6.
Atkinson AC
,
Yeh L
(1982)
Inference for Sichel’s compound Poisson distribution
. Journal of the American Statistical Association,
77,
153-158
.
7.
Breslow NE
,
Clayton DG
(1993)
Approximate inference in generalized linear models
. Journal of the American Statistical Association,
88,
9-25
.
8.
Booth JG
,
Hobert JP
(1999)
Maximizing generalized linear mixed model likelihoods with an automated
Monte Carlo EM algorithm
. Journal of the Royal Statistical Society, 61,
265-285
.
9.
Chen JJ
,
Ahn H
(1996)
Fitting mixed Poisson regression models using quasi-likelihood methods
. Biometrical Journal, 38,
81-96
.
10.
Dean CB
(1992)
Testing for overdispersion in Poisson and binomial regression models
. Journal of the American Statistical Association,
87,
451-457
.
11.
Dean CB
(1994)
Modified pseudo-likelihood estimator of the overdispersion parameter in
Poisson mixture models
. Journal of Applied Statistics, 21,
523-533
.
12.
Dean CB
,
Lawless J
,
Willmot GE
(1989)
A mixed Poisson-inverse Gaussian regression model
. Canadian Journal of Statistics, 17,
171-182
.
13.
Dempster AP
,
Laird NM
,
Rubin D
(1977)
Maximum likelihood from incomplete data via the EM algorithm
. Journal of the Royal Statistical Society, B39,
1-38
.
14.
Diebolt J
,
Ip EHS
(1996) Stochastic EM: method and application. In
Gilks WR
,
Richardson S
,
Spiegelhalter DJ
eds. Markov chain Monte Carlo in practice.
London: Chapman and Hall
, 259-273
15.
Gourieroux C
,
Monfort A
,
Trongon, A.
(1984)
Pseudo maximum likelihood methods: applications to Poisson models
. Econometrica, 52,
701-720
.
16.
Hinde JP
(1982) Compound Poisson regression models. In
Gilchrist R
, editor, GLIM ’82New York: Springer-Verlag
, 109-121.
17.
Hinde J
,
Demetrio CGB
(1998)
Overdispersion: models and estimation
. Computational Statistics and Data Analysis, 27,
151-170
.
18.
Jank W
,
Booth J
(2001) Efficiency of Monte Carlo EM and simulated maximum
likelihood in generalized linear mixed models. Technical Report 2001-001,
Department of Statistics, University of Florida
.
19.
Johnson N
,
Kotz S
,
Kemp AW
(1992) Univariate discrete distributions.
New York: Wiley
.
20.
Jorgensen B
(1982) The generalized Inverse-Gaussian distribution.
Lecture notes in statistics, 9,
Berlin: Springer-Verlag
.
21.
Kaas R
,
Hesselager O
(1995)
Ordering claim size distributions and mixed Poisson probabilities
. Insurance: Mathematics and Economics, 17,
193-201
.
22.
Karlis D (1998) Estimation and hypothesis testing problems in Poisson
mixtures. PhD Thesis, Department of Statistics, Athens University of Economics.
23.
Lawless J
(1987)
Negative binomial and mixed Poisson regression
. Canadian Journal of Statistics, 15,
209-225
.
24.
Lee Y
,
Nelder JA
(1996)
Hierarchical generalized linear models
. Journal of the Royal Statistical Society, B58,
619-678
.
25.
McLachlan G
,
Krishnan N
(1997) The EM algorithm and extensions.
New York: Wiley
.
26.
McNeney B
,
Petkau J
(1994)
Overdispersed Poisson regression models for studies of air pollution and
human health
. Canadian Journal of Statistics, 22,
421-440
.
27.
Meng XL
,
Rubin D
(1993)
Maximum likelihood estimation via the ECM algorithm: A general framework
. Biometrika, 80,
267-278
.
28.
Munkin MK
,
Trivedi PK
(1999)
Simulated maximum likelihood estimation of multivariate mixed Poisson
regression models, with application
. Econometrics Journal, 2,
29-48
.
29.
Sapatinas T
(1995)
Identifiability of mixtures of power series distributions and related characterizations
. Annals of the Institute of Statistical Mathematics,
47,
447-459
.
30.
Sichel HS
(1974)
On a distribution representing sentence-length in written prose
. Journal of the Royal Statistical Society, A137,
25-34
.
31.
Sichel HS
(1982)
Asymptotic efficiencies of three methods of estimation for the inverse
Gaussian-Poisson distribution
. Biometrika, 69,
467-472
.
32.
Shaban SA
(1988) Poisson-Lognormal Distributions. In
Crow EL
,
Shimizu K
eds. Log-normal distributions: theory and applications.
New York: Marcel Decker
, 195-210.
33.
Stein GZ
,
Yuritz JM
(1988)
Linear models with an Inverse-Gaussian Poisson error distribution
. Communication in Statistics - Theory and Methods,
17,
557-571
.
34.
Stein GZ
,
Zucchini W
,
Juritz JM
(1987)
Parameter estimation for the Sichel distribution and its multivariate extension
. Journal of the American Statistical Association,
82,
938-944
.
35.
Tanner MA
(1996) Tools for statistical inference: methods for the
exploration of posterior distributions and likelihood functions3rd edn.Berlin: Springer-Verlag
.
36.
Thall PF
(1988)
Mixed Poisson likelihood regression models for longitudinal interval count data
. Biometrics, 44,
197-209
.
37.
Van De Ven R
,
Weber NC
(1995)
Log-linear models for mean and dispersion in mixed Poisson regression models
. Australian Journal of Statistics, 37,
205-216
.
38.
Walker S
(1996)
An EM algorithm for nonlinear random effects models
. Biometrics, 52,
934-944
.
39.
Wang P
,
Puterman M
,
Cokburn I
,
Le N
(1996)
Mixed Poisson regression models with covariate dependent rates
. Biometrics, 52,
381-400
.
40.
Wei GCG
,
Tanner MA
(1990)
A Monte Carlo implementation of the EM algorithm and the poor
man’s data augmentation algorithms
. Journal of the American Statistical Association,
85,
699-704
.
41.
Willmot G
(1987)
The Poisson-inverse Gaussian distribution as an alternative to the negative binomial
. Scandinavian Actuarial Journal, 87,
113-127
.
42.
Willmot G
(1993)
On recursive evaluation of mixed Poisson probabilities and related quantities
. Scandinavian Actuarial Journal, 93,
101-113
.
43.
Xue D
,
Deddens JA
(1992)
Overdispersed negative binomial regression models
. Communications in Statistics-Theory and Methods,
21,
2215-2226
.
