Maximum Likelihood Estimation in a Latent Variable Problem by David R . Brillinger
نویسندگان
چکیده
منابع مشابه
A comparison of algorithms for maximum likelihood estimation of Spatial GLM models
In spatial generalized linear mixed models, spatial correlation is assumed by adding normal latent variables to the model. In these models because of the non-Gaussian spatial response and the presence of latent variables the likelihood function cannot usually be given in a closed form, thus the maximum likelihood approach is very challenging. The main purpose of this paper is to introduce two n...
متن کاملMaximum Likelihood Estimation of Parameters in Generalized Functional Linear Model
Sometimes, in practice, data are a function of another variable, which is called functional data. If the scalar response variable is categorical or discrete, and the covariates are functional, then a generalized functional linear model is used to analyze this type of data. In this paper, a truncated generalized functional linear model is studied and a maximum likelihood approach is used to esti...
متن کاملLearning Mixture Models with the Latent Maximum Entropy Principle
We present a new approach to estimating mixture models based on a new inference principle we have proposed: the latent maximum entropy principle (LME). LME is different both from Jaynes’ maximum entropy principle and from standard maximum likelihood estimation. We demonstrate the LME principle by deriving new algorithms for mixture model estimation, and show how robust new variants of the EM al...
متن کاملParticle methods for maximum likelihood estimation in latent variable models
Standard methods for maximum likelihood parameter estimation in latent variable models rely on the Expectation-Maximization algorithm and its Monte Carlo variants. Our approach is different and motivated by similar considerations to simulated annealing; that is we build a sequence of artificial distributions whose support concentrates itself on the set of maximum likelihood estimates. We sample...
متن کاملTwo Papers on Bilinear Systems : a ) A Study of Second - and Third - order Spectral Procedures and Maximum Likelihood
This paper describes a comparative investigation of two spectral momentbased identification procedures with each other and with the method of maximum likelihood. The case considered is that of a bilinear system. Analyses making use of secondand third-order moments are presented. Inputs investigated are pulse, pulsepair and stochastic. Some of the relative advantages and disadvantages of each es...
متن کامل