Monte Carlo EM algorithm for generalized linear models with linear structural random effects

Author

Zhou, Xingcai

Author_Institution

Dept. of Math. & Comput. Sci., Tongling Univ., Tongling, China

Volume

1

fYear

2011

fDate

11-13 March 2011

Firstpage

444

Lastpage

447

Abstract

We propose generalized linear mixed models with linear structural random effects based on generalized linear mixed models. In this article, an Monte Carlo EM type algorithm is developed for maximum likelihood estimations of the proposed models. To avoid computation of the complicated multiple integrals involved, the E-step is completed by Monte Carlo techniques. The M-step is carried out efficiently by simple Conditional Maximization. Standard errors of MLEs are obtained via Louis´s identity.

Keywords

Monte Carlo methods; computational complexity; maximum likelihood estimation; MLE; Monte Carlo EM algorithm; conditional maximization; generalized linear model; linear structural random effect; maximum likelihood estimation; Approximation algorithms; Biological system modeling; Computational modeling; Equations; Mathematical model; Maximum likelihood estimation; Monte Carlo methods; EM algorithm; Linear structural; Monte Carlo; Standard errors estimates; generalized linear mixed models;

fLanguage

English

Publisher

ieee

Conference_Titel

Computer Research and Development (ICCRD), 2011 3rd International Conference on

Conference_Location

Shanghai

Print_ISBN

978-1-61284-839-6

Type

conf

DOI

10.1109/ICCRD.2011.5764054

Filename

5764054