Estimation and hypothesis testing in mixed linear models

Dissertação apresentada para obtenção do Grau de Doutor em Matemática, Estatística, pela Universidade Nova de Lisboa, faculdade de Ciências e Tecnologia

Improved testing inference in mixed linear models

Mixed linear models are commonly used in repeated measures studies. They account for the dependence amongst observations obtained from the same experimental unit. Often, the number of observations is small, and it is thus important to use inference strategies that incorporate small sample corrections. In this paper, we develop modified versions of the likelihood ratio test for fixed effects inference in mixed linear models. In particular, we derive a Bartlett correction to such a test, and also to a test obtained from a modified profile likelihood function. Our results generalize those in [Zucker, D.M., Lieberman, O., Manor, O., 2000. Improved small sample inference in the mixed linear model: Bartlett correction and adjusted likelihood. Journal of the Royal Statistical Society B, 62,827-838] by allowing the parameter of interest to be vector-valued. Additionally, our Bartlett corrections allow for random effects nonlinear covariance matrix structure. We report simulation results which show that the proposed tests display superior finite sample behavior relative to the standard likelihood ratio test. An application is also presented and discussed. (C) 2008 Elsevier B.V. All rights reserved.

Local Distance-Based Generalized Linear Models using the dbstats package for R

This paper introduces local distance-based generalized linear models. These models extend (weighted) distance-based linear models firstly with the generalized linear model concept, then by localizing. Distances between individuals are the only predictor information needed to fit these models. Therefore they are applicable to mixed (qualitative and quantitative) explanatory variables or when the regressor is of functional type. Models can be fitted and analysed with the R package dbstats, which implements several distancebased prediction methods.

Local influence for Student-t partially linear models

In this paper we extend partial linear models with normal errors to Student-t errors Penalized likelihood equations are applied to derive the maximum likelihood estimates which appear to be robust against outlying observations in the sense of the Mahalanobis distance In order to study the sensitivity of the penalized estimates under some usual perturbation schemes in the model or data the local influence curvatures are derived and some diagnostic graphics are proposed A motivating example preliminary analyzed under normal errors is reanalyzed under Student-t errors The local influence approach is used to compare the sensitivity of the model estimates (C) 2010 Elsevier B V All rights reserved

Hglm: A package for fitting hierarchical generalized linear models

We present the hglm package for fitting hierarchical generalized linear models. It can be used for linear mixed models and generalized linear mixed models with random effects for a variety of links and a variety of distributions for both the outcomes and the random effects. Fixed effects can also be fitted in the dispersion part of the model.

Genetic heterogeneity of residual variance : estimation of variance components using double hierarchical generalized linear models

Background: The sensitivity to microenvironmental changes varies among animals and may be under genetic control. It is essential to take this element into account when aiming at breeding robust farm animals. Here, linear mixed models with genetic effects in the residual variance part of the model can be used. Such models have previously been fitted using EM and MCMC algorithms. Results: We propose the use of double hierarchical generalized linear models (DHGLM), where the squared residuals are assumed to be gamma distributed and the residual variance is fitted using a generalized linear model. The algorithm iterates between two sets of mixed model equations, one on the level of observations and one on the level of variances. The method was validated using simulations and also by re-analyzing a data set on pig litter size that was previously analyzed using a Bayesian approach. The pig litter size data contained 10,060 records from 4,149 sows. The DHGLM was implemented using the ASReml software and the algorithm converged within three minutes on a Linux server. The estimates were similar to those previously obtained using Bayesian methodology, especially the variance components in the residual variance part of the model. Conclusions: We have shown that variance components in the residual variance part of a linear mixed model can be estimated using a DHGLM approach. The method enables analyses of animal models with large numbers of observations. An important future development of the DHGLM methodology is to include the genetic correlation between the random effects in the mean and residual variance parts of the model as a parameter of the DHGLM.

Influence diagnostics for elliptical semiparametric mixed models

In this paper we extend semiparametric mixed linear models with normal errors to elliptical errors in order to permit distributions with heavier and lighter tails than the normal ones. Penalized likelihood equations are applied to derive the maximum penalized likelihood estimates (MPLEs) which appear to be robust against outlying observations in the sense of the Mahalanobis distance. A reweighed iterative process based on the back-fitting method is proposed for the parameter estimation and the local influence curvatures are derived under some usual perturbation schemes to study the sensitivity of the MPLEs. Two motivating examples preliminarily analyzed under normal errors are reanalyzed considering some appropriate elliptical errors. The local influence approach is used to compare the sensitivity of the model estimates.

LIKELIHOOD ESTIMATION OF CONJUGACY RELATIONSHIPS IN LINEAR MODELS WITH APPLICATIONS TO HIGH-THROUGHPUT GENOMICS

In the simultaneous estimation of a large number of related quantities, multilevel models provide a formal mechanism for efficiently making use of the ensemble of information for deriving individual estimates. In this article we investigate the ability of the likelihood to identify the relationship between signal and noise in multilevel linear mixed models. Specifically, we consider the ability of the likelihood to diagnose conjugacy or independence between the signals and noises. Our work was motivated by the analysis of data from high-throughput experiments in genomics. The proposed model leads to a more flexible family. However, we further demonstrate that adequately capitalizing on the benefits of a well fitting fully-specified likelihood in the terms of gene ranking is difficult.

Linear models of simple cells: Correspondence to real cell responses and space spanning properties

Despite their limitations, linear filter models continue to be used to simulate the receptive field properties of cortical simple cells. For theoreticians interested in large scale models of visual cortex, a family of self-similar filters represents a convenient way in which to characterise simple cells in one basic model. This paper reviews research on the suitability of such models, and goes on to advance biologically motivated reasons for adopting a particular group of models in preference to all others. In particular, the paper describes why the Gabor model, so often used in network simulations, should be dropped in favour of a Cauchy model, both on the grounds of frequency response and mutual filter orthogonality.

Modelos Mixtos-Aditivos Generalizados para el Estudio de Fenómenos Biológicos y Sociales Descriptos por Datos Agregados y/o Longitudinales. Generalized Mixed Additive Models for the Study of Biological and Social Phenomenons from Longitudinal and/or Aggregated Data.

Este proyecto propone extender y generalizar los procesos de estimación e inferencia de modelos aditivos generalizados multivariados para variables aleatorias no gaussianas, que describen comportamientos de fenómenos biológicos y sociales y cuyas representaciones originan series longitudinales y datos agregados (clusters). Se genera teniendo como objeto para las aplicaciones inmediatas, el desarrollo de metodología de modelación para la comprensión de procesos biológicos, ambientales y sociales de las áreas de Salud y las Ciencias Sociales, la condicionan la presencia de fenómenos específicos, como el de las enfermedades.Es así que el plan que se propone intenta estrechar la relación entre la Matemática Aplicada, desde un enfoque bajo incertidumbre y las Ciencias Biológicas y Sociales, en general, generando nuevas herramientas para poder analizar y explicar muchos problemas sobre los cuales tienen cada vez mas información experimental y/o observacional.Se propone, en forma secuencial, comenzando por variables aleatorias discretas (Yi, con función de varianza menor que una potencia par del valor esperado E(Y)) generar una clase unificada de modelos aditivos (paramétricos y no paramétricos) generalizados, la cual contenga como casos particulares a los modelos lineales generalizados, no lineales generalizados, los aditivos generalizados, los de media marginales generalizados (enfoques GEE1 -Liang y Zeger, 1986- y GEE2 -Zhao y Prentice, 1990; Zeger y Qaqish, 1992; Yan y Fine, 2004), iniciando una conexión con los modelos lineales mixtos generalizados para variables latentes (GLLAMM, Skrondal y Rabe-Hesketh, 2004), partiendo de estructuras de datos correlacionados. Esto permitirá definir distribuciones condicionales de las respuestas, dadas las covariables y las variables latentes y estimar ecuaciones estructurales para las VL, incluyendo regresiones de VL sobre las covariables y regresiones de VL sobre otras VL y modelos específicos para considerar jerarquías de variación ya reconocidas. Cómo definir modelos que consideren estructuras espaciales o temporales, de manera tal que permitan la presencia de factores jerárquicos, fijos o aleatorios, medidos con error como es el caso de las situaciones que se presentan en las Ciencias Sociales y en Epidemiología, es un desafío a nivel estadístico. Se proyecta esa forma secuencial para la construcción de metodología tanto de estimación como de inferencia, comenzando con variables aleatorias Poisson y Bernoulli, incluyendo los existentes MLG, hasta los actuales modelos generalizados jerárquicos, conextando con los GLLAMM, partiendo de estructuras de datos correlacionados. Esta familia de modelos se generará para estructuras de variables/vectores, covariables y componentes aleatorios jerárquicos que describan fenómenos de las Ciencias Sociales y la Epidemiología.

Approximations of the Fisher information for the construction of efficient experimental designs in nonlinear mixed effects models

Magdeburg, Univ., Fak. für Mathematik, Diss., 2012

Optimal designs for multivariate linear models

Magdeburg, Univ., Fak. für Mathematik, Diss., 2013