GLIMMIX: Software for estimating mixtures and mixtures of generalized linear models

GLIMMIX is a commercial WINDOWS-based computer program that implements the EM algorithm (Dempster, Laird and Rubin 1977) for the estimation of finite mixtures and mixtures of generalized linear models. The program allows for the specification of a number of distributions in the exponential family, i ncluding the normal, gamma, binomial, Poisson, and multinomial distribution s. For each of those distributions, a variety of link functions can be spec ified to relate the expectation of the dependent variable to a linear predi ctor. Several statistics, including AIC, CAIC and BIC are computed to aid i n model selection (cf. Akaike 1974; Bozdogan 1987), missing values are acco mmodated, and posterior membership probabilities are computed For cases, in cluded or not included in the analysis. Simple discriminant type models dea ling with concomitant variables to describe the classes are supported, and a random responder class can be added to the model. Various graphs are prov ided. A demonstration version of the program can be obtained from http://ww w/gamma.rug.nl. Before providing some details on the GLIMMIX software, a br ief review of a few relevant issues in Mixture modelling are provided.