Pointwise and functional approximations in Monte Carlo maximum likelihood estimation

We consider the use of Monte Carlo methods to obtain maximum likelihood estimates for random effects models and distinguish between the pointwise and functional approaches. We explore the relationship between the two approaches and compare them with the EM algorithm. The functional approach is more ambitious but the approximation is local in nature which we demonstrate graphically using two simple examples. A remedy is to obtain successively better approximations of the relative likelihood function near the true maximum likelihood estimate. To save computing time, we use only one Newton iteration to approximate the maximiser of each Monte Carlo likelihood and show that this is equivalent to the pointwise approach. The procedure is applied to fit a latent process model to a set of polio incidence data. The paper ends by a comparison between the marginal likelihood and the recently proposed hierarchical likelihood which avoids integration altogether.

[1]  K. Chan,et al.  Monte Carlo EM Estimation for Time Series Models Involving Counts , 1995 .

[2]  The Evaluation of Integrals , 1984 .

[3]  D. M. Smith.  Algorithm AS 189: Maximum Likelihood Estimation of the Parameters of the Beta Binomial Distribution , 1983 .

[4]  M. Karim.  Generalized Linear Models With Random Effects , 1991 .

[5]  S L Zeger,et al.  Generalized linear models with random effects; salamander mating revisited. , 1992, Biometrics.

[6]  E. Crouch,et al.  The Evaluation of Integrals of the form ∫+∞ −∞ f(t)exp(−t 2) dt: Application to Logistic-Normal Models , 1990 .

[7]  C. Geyer,et al.  Constrained Monte Carlo Maximum Likelihood for Dependent Data , 1992 .

[8]  K. Lange.  A gradient algorithm locally equivalent to the EM algorithm , 1995 .

[9]  C. A. McGilchrist,et al.  Restricted BLUP for mixed linear models , 1991 .

[10]  R. Schall.  Estimation in generalized linear models with random effects , 1991 .

[11]  D. Rubin,et al.  Maximum likelihood from incomplete data via the EM - algorithm plus discussions on the paper , 1977 .

[12]  N. Breslow,et al.  Approximate inference in generalized linear mixed models , 1993 .

[13]  C. S. Weil.  Selection of the valid number of sampling units and a consideration of their combination in toxicological studies involving reproduction, teratogenesis or carcinogenesis. , 1970, Food and cosmetics toxicology.

[14]  C. Mcgilchrist,et al.  Regression with frailty in survival analysis. , 1991, Biometrics.

[15]  C. Mcgilchrist.  Estimation in Generalized Mixed Models , 1994 .

[16]  A. Kuk,et al.  The monte carlo newton-raphson algorithm , 1997 .

[17]  T. Louis.  Finding the Observed Information Matrix When Using the EM Algorithm , 1982 .

[18]  Murray Aitkin,et al.  Variance Component Models with Binary Response: Interviewer Variability , 1985 .

[19]  C. McCulloch.  Maximum Likelihood Algorithms for Generalized Linear Mixed Models , 1997 .

[20]  Joel C. Kleinman,et al.  Proportions with Extraneous Variance: Single and Independent Samples , 1973 .

[21]  J. Nelder,et al.  Hierarchical Generalized Linear Models , 1996 .

[22]  S. Zeger.  A regression model for time series of counts , 1988 .

[23]  C. Geyer.  On the Convergence of Monte Carlo Maximum Likelihood Calculations , 1994 .

[24]  G. C. Wei,et al.  A Monte Carlo Implementation of the EM Algorithm and the Poor Man's Data Augmentation Algorithms , 1990 .

[25]  McGilchrist Ca,et al.  Regression with frailty in survival analysis. , 1991 .