Bayesian Estimation of the Logistic Positive Exponent IRT Model

A Bayesian inference approach using Markov Chain Monte Carlo (MCMC) is developed for the logistic positive exponent (LPE) model proposed by Samejima and for a new skewed Logistic Item Response Theory (IRT) model, named Reflection LPE model. Both models lead to asymmetric item characteristic curves (ICC) and can be appropriate because a symmetric ICC treats both correct and incorrect answers symmetrically, which results in a logical contradiction in ordering examinees on the ability scale. A data set corresponding to a mathematical test applied in Peruvian public schools is analyzed, where comparisons with other parametric IRT models also are conducted. Several model comparison criteria are discussed and implemented. The main conclusion is that the LPE and RLPE IRT models are easy to implement and seem to provide the best fit to the data set considered.

Framework for Skew-Probit Links in Binary Regression

We review several asymmetrical links for binary regression models and present a unified approach for two skew-probit links proposed in the literature. Moreover, under skew-probit link, conditions for the existence of the ML estimators and the posterior distribution under improper priors are established. The framework proposed here considers two sets of latent variables which are helpful to implement the Bayesian MCMC approach. A simulation study to criteria for models comparison is conducted and two applications are made. Using different Bayesian criteria we show that, for these data sets, the skew-probit links are better than alternative links proposed in the literature.

A Bayesian Semiparametric Approach for Solving the Discrete Calibration Problem

In this article, we introduce a semi-parametric Bayesian approach based on Dirichlet process priors for the discrete calibration problem in binomial regression models. An interesting topic is the dosimetry problem related to the dose-response model. A hierarchical formulation is provided so that a Markov chain Monte Carlo approach is developed. The methodology is applied to simulated and real data.

Likelihood ratio tests for variance components in linear mixed models

Although the asymptotic distributions of the likelihood ratio for testing hypotheses of null variance components in linear mixed models derived by Stram and Lee [1994. Variance components testing in longitudinal mixed effects model. Biometrics 50, 1171-1177] are valid, their proof is based on the work of Self and Liang [1987. Asymptotic properties of maximum likelihood estimators and likelihood tests under nonstandard conditions. J. Amer. Statist. Assoc. 82, 605-610] which requires identically distributed random variables, an assumption not always valid in longitudinal data problems. We use the less restrictive results of Vu and Zhou [1997. Generalization of likelihood ratio tests under nonstandard conditions. Ann. Statist. 25, 897-916] to prove that the proposed mixture of chi-squared distributions is the actual asymptotic distribution of such likelihood ratios used as test statistics for null variance components in models with one or two random effects. We also consider a limited simulation study to evaluate the appropriateness of the asymptotic distribution of such likelihood ratios in moderately sized samples. (C) 2008 Elsevier B.V. All rights reserved.

Asymptotic skewness in Birnbaum-Saunders nonlinear regression models

The family of distributions proposed by Birnbaum and Saunders (1969) can be used to model lifetime data and it is widely applicable to model failure times of fatiguing materials. We give a simple matrix formula of order n(-1/2), where n is the sample size, for the skewness of the distributions of the maximum likelihood estimates of the parameters in Birnbaum-Saunders nonlinear regression models, recently introduced by Lemonte and Cordeiro (2009). The formula is quite suitable for computer implementation, since it involves only simple operations on matrices and vectors, in order to obtain closed-form skewness in a wide range of nonlinear regression models. Empirical and real applications are analyzed and discussed. (C) 2010 Elsevier B.V. All rights reserved.

Bayesian Analysis for the Generalized Lognormal Distribution Applied to Failure Time Analysis

There are several versions of the lognormal distribution in the statistical literature, one is based in the exponential transformation of generalized normal distribution (GN). This paper presents the Bayesian analysis for the generalized lognormal distribution (logGN) considering independent non-informative Jeffreys distributions for the parameters as well as the procedure for implementing the Gibbs sampler to obtain the posterior distributions of parameters. The results are used to analyze failure time models with right-censored and uncensored data. The proposed method is illustrated using actual failure time data of computers.

Predictive models for chronic renal disease using decision trees, naïve bayes and case-based methods

Data mining can be used in healthcare industry to “mine” clinical data to discover hidden information for intelligent and affective decision making. Discovery of hidden patterns and relationships often goes intact, yet advanced data mining techniques can be helpful as remedy to this scenario. This thesis mainly deals with Intelligent Prediction of Chronic Renal Disease (IPCRD). Data covers blood, urine test, and external symptoms applied to predict chronic renal disease. Data from the database is initially transformed to Weka (3.6) and Chi-Square method is used for features section. After normalizing data, three classifiers were applied and efficiency of output is evaluated. Mainly, three classifiers are analyzed: Decision Tree, Naïve Bayes, K-Nearest Neighbour algorithm. Results show that each technique has its unique strength in realizing the objectives of the defined mining goals. Efficiency of Decision Tree and KNN was almost same but Naïve Bayes proved a comparative edge over others. Further sensitivity and specificity tests are used as statistical measures to examine the performance of a binary classification. Sensitivity (also called recall rate in some fields) measures the proportion of actual positives which are correctly identified while Specificity measures the proportion of negatives which are correctly identified. CRISP-DM methodology is applied to build the mining models. It consists of six major phases: business understanding, data understanding, data preparation, modeling, evaluation, and deployment.

Parameter estimation of Poisson generalized linear mixed models based on three different statistical principles : a simulation study

Generalized linear mixed models are flexible tools for modeling non-normal data and are useful for accommodating overdispersion in Poisson regression models with random effects. Their main difficulty resides in the parameter estimation because there is no analytic solution for the maximization of the marginal likelihood. Many methods have been proposed for this purpose and many of them are implemented in software packages. The purpose of this study is to compare the performance of three different statistical principles - marginal likelihood, extended likelihood, Bayesian analysis-via simulation studies. Real data on contact wrestling are used for illustration.

Bayesian analysis of random regression models using B-splines to model test-day milk yield of Holstein cattle in Brazil

Conselho Nacional de Desenvolvimento Científico e Tecnológico (CNPq)

Stress Analysis in Simulation Models With or Without Implant Threads Representation

Purpose: This study aimed to evaluate the influence of implants with or without threads representation on the outcome of a two-dimensional finite element (FE) analysis. Materials and Methods: Two-dimensional FE models that reproduced a frontal section of edentulous mandibular posterior bone were constructed using a standard crown/implant/screw system representation. To evaluate the effect of implant threads, two models were created: a model in which the implant threads were accurately simulated (precise model) and a model in which implants with a smooth surface (press-fit implant) were used (simplified model). An evaluation was performed on ANSYS software, in which a load of 133 N was applied at a 30-degree angulation and 2 mm off-axis from the long axis of the implant on the models, The Von Mises stresses were measured. Results: The precise model (1.45 MPa) showed higher maximum stress values than the simplified model (1.2 MPa). Whereas in the cortical bone, the stress values differed by about 36% (292.95 MPa for the precise model and 401.14 MPa for the simplified model), in trabecular bone (19.35 MPa and 20.35 MPa, respectively), the stress distribution and stress values were similar. Stress concentrations occurred around the implant neck and the implant apex. Conclusions: Considering implant and cortical bone analysis, remarkable differences in stress values were found between the models. Although the models showed different absolute stress values, the stress distribution was similar. INT J ORAL MAXILLOFAC IMPLANTS 2009;24:1040-1044

Methods to verify parameter equality in nonlinear regression models

Coordenação de Aperfeiçoamento de Pessoal de Nível Superior (CAPES)

Modelos alternativos para detecção de locos de características quantitativas (QTL) de carcaça e crescimento nos cromossomos 4, 5 e 7 de suínos

O conhecimento do genoma pode auxiliar na identificação de regiões cromossômicas e, eventualmente, de genes que controlam características quantitativas (QTLs) de importância econômica. em um experimento com 1.129 suínos resultantes do cruzamento entre machos da raça Meishan e fêmeas Large White e Landrace, foram analisadas as características gordura intramuscular (GIM), em %, e ganho dos 25 aos 90 kg de peso vivo (GP), em g/dia, em 298 animais F1 e 831 F2, e espessura de toucinho (ET), em mm, em 324 F1 e 805 F2. Os animais das gerações F1 e F2 foram tipificados com 29 marcadores microsatélites. Estudou-se a ligação entre os cromossomos 4, 6 e 7 com GIM, ET e GP. Análises de QTL utilizando-se metodologia Bayesiana foram aplicadas mediante três modelos genéticos: modelo poligênico infinitesimal (MPI); modelo poligênico finito (MPF), considerando-se três locos; e MPF combinado com MPI. O número de QTLs, suas respectivas posições nos três cromossomos e o efeito fenotípico foram estimados simultaneamente. Os sumários dos parâmetros estimados foram baseados nas distribuições marginais a posteriori, obtidas por meio do uso da Cadeia de Markov, algoritmos de Monte Carlo (MCMC). Foi possível evidenciar dois QTLs relacionados a GIM nos cromossomos 4 e 6 e dois a ET nos cromossomos 4 e 7. Somente quando se ajustou o MPI, foram observados QTLs no cromossomo 4 para ET e GIM. Não foi possível detectar QTLs para a característica GP com a aplicação dessa metodologia, o que pode ter resultado do uso de marcadores não informativos ou da ausência de QTLs segregando nos cromossomos 4, 6 e 7 desta população. Foi evidenciada a vantagem de se analisar dados experimentais ajustando diferentes modelos genéticos; essas análises ilustram a utilidade e ampla aplicabilidade do método Bayesiano.