Bayesian outlier analysis in binary regression

We propose alternative approaches to analyze residuals in binary regression models based on random effect components. Our preferred model does not depend upon any tuning parameter, being completely automatic. Although the focus is mainly on accommodation of outliers, the proposed methodology is also able to detect them. Our approach consists of evaluating the posterior distribution of random effects included in the linear predictor. The evaluation of the posterior distributions of interest involves cumbersome integration, which is easily dealt with through stochastic simulation methods. We also discuss different specifications of prior distributions for the random effects. The potential of these strategies is compared in a real data set. The main finding is that the inclusion of extra variability accommodates the outliers, improving the adjustment of the model substantially, besides correctly indicating the possible outliers.

Impacto das condições climáticas sobre a etiologia das bacteremias nosocomiais no Hospital das Clínicas da Faculdade de Medicina de Botucatu-UNESP

Pós-graduação em Doenças Tropicais - FMB

Stochastic simulation of time-series models combined with geostatistics to predict water-table scenarios in a Guarani Aquifer System outcrop area, Brazil

Fundação de Amparo à Pesquisa do Estado de São Paulo (FAPESP)

Estimation of genetic parameters for partial egg production periods by means of random regression models

Conselho Nacional de Desenvolvimento Científico e Tecnológico (CNPq)

Morphological Aspects and Storage of Dypsis leptocheilos (Hodel) Beentje & J. Dransf. (Arecaceae) Seeds

Dypsis leptocheilos (Hodel) Beentje & J. Dransf. palms are widely used in landscaping, however there is a lack of studies about its biology, especially on the morphology of its seedling disseminules and the ideal seed storage time without loss of viability. The objectives of this work were to describe the morphological aspects and to study the storage effect on the germination of D. leptocheilos seeds. The experiments were performed at FCAV/UNESP, Jaboticabal Campus, São Paulo State, Brazil. In order to obtain a morphology description 100 disseminules were sowed in vermiculite. The external and internal sides of the disseminules as well as the representative phases of the germination process were sketched with the help of a light chamber attached to a stereomicroscope. Five treatments were arranged in order to study the storage effects on the seeds (four months of storage and straight after harvest), with four replicates with 25 disseminules each (seeds with attached endocarp) in a completely randomized design. The germination test was conducted over a 140 days period and noted every two days. In order to calculate the germination percentage (% G) and the Germination Speed Index (GSI), the seeds were considered germinated when the germinative button appeared. The data were submitted to a polynomial regression analysis to verify the behavior of the variables over the storage period. The D. leptocheilos seeds have rounded shape, ruminate endosperm and hard consistency. The embryo is lateral, peripheral and relatively undifferentiated, with approximately 3 mm long and conical shape. The germination percentage increased from 17% for seeds sowed straight after harvest to 73% for those stored over a three-month period. For the GSI the increase was about 0.85%. It was concluded that the germination of D. leptocheilos is of an adjacent ligulate type. Furthermore, the seeds stored for 3 months at 14 degrees C showed both higher percentage and germination speed.

Parametric correlation functions to model the structure of permanent environmental (co)variances in milk yield random regression models

Fundação de Amparo à Pesquisa do Estado de São Paulo (FAPESP)

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)

Estimates of genetic parameters for scrotal circumference using random regression models in Nelore cattle

A total of 15,901 scrotal circumference (SC) records from 5300 Nelore bulls, ranging from 229 to 560 days of age, were used with the objective of estimating (co)variance functions for SC, using random regression models. Models included the fixed effects of contemporary group and age of dam at calving as covariable (linear and quadratic effects). To model the population mean trend, a third order Legendre polynomial on animal age was utilized. The direct additive genetic and animal permanent environmental random effects were modeled by Legendre polynomials on animal age, with orders of fit ranging from 1 to 5. Residual variances were modeled considering 1 (homogeneity of variance) or 4 age classes. Results obtained with the random regression models were compared to multi-trait analysis. (Co)variance estimates using multi-trait and random regression models were similar. The model considering a third- and fifth-order Legendre polynomials for additive genetic and animal permanent environmental effects, respectively, was the most adequate to model changes in variance of SC with age. Heritability estimates for SC ranged from 0.24 (229 days of age) to 0.47 (300 days of age), remained almost constant until 500 days of age (0.52), decreasing thereafter (0.44). In general, the genetic correlations between measures of scrotal circumference obtained from 229 to 560 days of age decreased with increasing distance between ages. For genetic evaluation scrotal circumference could be measured between 400 and 500 days of age. (C) 2010 Elsevier B.V. All rights reserved.

Random regression analyses using B-splines functions to model growth from birth to adult age in Canchim cattle

Conselho Nacional de Desenvolvimento Científico e Tecnológico (CNPq)

Random regression models to estimate genetic parameters for test-day milk yield in Brazilian Murrah buffaloes

Coordenação de Aperfeiçoamento de Pessoal de Nível Superior (CAPES)

Random regression models to estimate test-day milk yield genetic parameters Holstein cows in Southeastern Brazil

Fundação de Amparo à Pesquisa do Estado de São Paulo (FAPESP)

Modeling of average growth curve in Santa Ines sheep using random regression models

Conselho Nacional de Desenvolvimento Científico e Tecnológico (CNPq)

Use of ridge regression for the prediction of early growth performance in crossbred calves

Fundação de Amparo à Pesquisa do Estado de São Paulo (FAPESP)

Variance component estimates applying random regression models for test-day milk yield in Caracu heifers (Bos taurus Artiodactyla, Bovidae)

Conselho Nacional de Desenvolvimento Científico e Tecnológico (CNPq)

Breeding value accuracy estimates for growth traits using random regression and multi-trait models in Nelore cattle

Fundação de Amparo à Pesquisa do Estado de São Paulo (FAPESP)