Rosangela H. Loschi

Incipient fault detection in induction machine stator-winding using a fuzzy-Bayesian change point detection approach

In this paper the incipient fault detection problem in induction machine stator-winding is considered. The problem is solved using a new technique of change point detection in time series, based on a three-step formulation. The technique can detect up to two change points in the time series. The first step consists of a Kohonen neural network classification algorithm that defines the model to be used, one change point or two change points. The second step consists of a fuzzy clustering to transform the initial data, with arbitrary distribution, into a new one that can be approximated by a beta distribution. The fuzzy cluster centers are determined by using the Kohonen neural network classification algorithm used in the first step. The last step consists in using the Metropolis-Hastings algorithm for performing the change point detection in the transformed time series generated by the second step with that known distribution. The incipient faults are detected as long as they characterize change points in such transformed time series. The main contribution of the proposed approach in this paper, related to previous one in the Literature, is to detect up to two change points in the time series considered, besides the enhanced resilience of the new fault detection procedure against false alarms, combined with a good sensitivity that allows the detection of rather small fault signals. Simulation results are presented to illustrate the proposed methodology.

Uso da autópsia verbal na investigação de óbitos com causa mal definida em Minas Gerais, Brasil

Ill-defined causes of death can be related to problems in access to health services or poor quality of medical care and are indicators of data quality in the Mortality Information System (MIS). A sample of municipalities (counties) was selected from the Northeastern Macro-Region of Minas Gerais State, Brazil, with the aim of investigating deaths from ill-defined causes and deaths not reported to the Mortality Information System in 2007, using the verbal autopsy technique. The method allowed identifying 87% of the causes of investigated deaths, of which 17% (n = 37) were due to violent causes. At the end of the study, of the 779 investigated deaths, 9.5% (n = 74) were due to external causes found outside the MIS. The distribution of causes was similar when comparing deaths reported (versus not reported) to the MIS for natural causes, but different when external causes were included. The article concludes that the verbal autopsy method can be a valuable tool for improving the MIS, allowing the identification of causes of death and improving data completeness.

Extension to the product partition model: computing the probability of a change

The well-known product partition model (PPM) is considered for the identification of multiple change points in the means and variances of normal data sequences. In a natural fashion, the PPM may provide product estimates of these parameters at each instant of time, as well as the posterior distributions of the partitions and the number of change points. Prior distributions are assumed for the means, variances, and for the probability p that each individual time is a change point. The PPM is extended to generate the posterior distribution of p and the posterior probability that each instant of time is a change point. A Gibbs sampling scheme is used to compute all estimates of interest. The methodology is applied to an important time series from the Brazilian stock market. A sensitivity analysis is performed assuming different prior specifications of p.

A Gibbs sampling scheme to the product partition model: an application to change-point problems

This paper extends previous results for the classical product partition model applied to the identification of multiple change points in the means and variances of time series. Prior distributions for these two parameters and for the probability p that a change takes place at a particular period of time are considered and a new scheme based on Gibbs sampling to estimate the posterior relevances of the model is proposed. The resulting algorithm is applied to the analysis of two Brazilian stock market data. The computational experiments seem to indicate that the algorithm runs fast in common PC-like machines and it may be a useful tool for analyzing change-point problems.

Shape mixtures of multivariate skew-normal distributions

Classes of shape mixtures of independent and dependent multivariate skew-normal distributions are considered and some of their main properties are studied. If interpreted from a Bayesian point of view, the results obtained in this paper bring tractability to the problem of inference for the shape parameter, that is, the posterior distribution can be written in analytic form. Robust inference for location and scale parameters is also obtained under particular conditions.

Estimating the grid of time-points for the piecewise exponential model

One of the greatest challenges related to the use of piecewise exponential models (PEMs) is to find an adequate grid of time-points needed in its construction. In general, the number of intervals in such a grid and the position of their endpoints are ad-hoc choices. We extend previous works by introducing a full Bayesian approach for the piecewise exponential model in which the grid of time-points (and, consequently, the endpoints and the number of intervals) is random. We estimate the failure rates using the proposed procedure and compare the results with the non-parametric piecewise exponential estimates. Estimates for the survival function using the most probable partition are compared with the Kaplan–Meier estimators (KMEs). A sensitivity analysis for the proposed model is provided considering different prior specifications for the failure rates and for the grid. We also evaluate the effect of different percentage of censoring observations in the estimates. An application to a real data set is also provided. We notice that the posteriors are strongly influenced by prior specifications, mainly for the failure rates parameters. Thus, the priors must be fairly built, say, really disclosing the expert prior opinion.

An analysis of the influence of some prior specifications in the identification of change points via product partition model

In this paper, we consider the product partition model for the estimation of normal means and variances of a sequence of observations that experiences changes in these parameters at unknown times. The estimates of the parameters by using product partition model are the weighted average of the estimates based in blocks (groups) of observations by the posterior relevance of these blocks which depends on the prior cohesions. We implement the Barry and Hartigans method to this change point problem and propose an easy-to-implement modification to their method. We use Yaos prior cohesions and investigate the influence of different prior distributions to the parameter that indexes these cohesions in the product estimates. A comparison between the estimates obtained by using both these methods and those provided by using Yaos method is done considering different settings for its application. We apply the three methods presented in this paper to stock market data. The results seem to indicate that the method proposed is competitive and also that the prior specifications influence in the product estimates.

Mortalidade neonatal precoce hospitalar em Minas Gerais: associação com variáveis assistenciais e a questão da subnotificação

INTRODUCAO: Os obitos neonatais precoces estao relacionados com problemas de acesso a assistencia de qualidade ao pre-natal, ao parto hospitalar e ao recem-nascido. Os hospitais em Minas Gerais estao distribuidos de forma heterogenea e isto pode se refletir em diferentes niveis de mortalidade neonatal (MN) entre as regioes do Estado. OBJETIVO: Investigar a MN precoce hospitalar no Sistema de Informacoes Hospitalares (SIH/SUS) e avaliar possivel associacao da taxa de mortalidade neonatal precoce obtida a partir do SIH/SUS (TMNP_SIH), com variaveis relativas ao atendimento a gestante e ao recem-nascido em estratos de municipios homogeneos. METODO: Utilizou-se o SIH/SUS para obter o numero de nascimentos e obitos segundo o municipio de residencia. Os municipios foram agrupados segundo microrregiao e tamanho populacional, totalizando 199 grupamentos. O metodo CART (Classification and Regression Tree) identificou tres estratos de grupamentos de municipios homogeneos do ponto de vista socioeconomico. Para cada estrato utilizou-se a matriz de correlacao de Spearman para avaliar associacao entre a TMNP e indicadores da assistencia. RESULTADOS: A TMNP_SIH para Minas Gerais foi de 10,9/1000 nascidos vivos. Observou-se maior probabilidade de TMNP_SIH menor que 8/1000 nascidos vivos onde a situacao socioeconomica e mais precaria. Observou-se correlacao positiva entre TMNP_SIH e bercos por mulher em idade fertil e baixo peso ao nascer no Estrato 1, e entre TMNP_SIH e medicos por habitante no Estrato 3. CONCLUSAO: Questoes relativas ao acesso a assistencia, sub-registro de obitos e erro de classificacao de neomorto como natimorto podem estar se refletindo na TMNP_SIH. O volume de nascimentos e obitos registrados no SIH/SUS justifica investimentos na qualidade desses registros e sua utilizacao em estudos epidemiologicos.

Parameter Interpretation in Skewed Logistic Regression with Random Intercept

This paper aims at providing the prior and posterior interpretations for the parameters in the logistic regression model with random or cluster-level intercept when univariate and multivariate classes of skew normal distributions are assumed to model the random effects behavior. We obtain the prior distributions for the odds ratio and their medians under skew normality for the random effects. Original results related to linear combinations of skew-normal distributions are obtained as a by-product and, in the univariate case, a new class of log-skew-normal distribution is introduced. Robust results are obtained whenever a class of multivariate skew-normal distribution is assumed. We also evaluate the effect of the misspecification of the random effects distributions in the odds ratio estimation. We consider both simulated and the Teratogenic activity experiment datasets. The latter was previously analysed in the literature. We concluded that the misspecification of the random effects distribution yields poor odds ratios estimates and that the median odds ratio is not necessarily the best measure of heterogeneity among the clusters as suggested in the literature.

Predictivistic characterizations of multivariate student- t models

De Finetti style theorems characterize models (predictive distributions) as mixtures of the likelihood function and the prior distribution, beginning from some judgment of invariance about observable quantities. The likelihood function generally has its functional form identified from invariance assumptions only. However, we need additional conditions on observable quantities (typically, assumptions on conditional expectations) to identify the prior distribution. In this paper, we consider some well-known invariance assumptions and establish additional conditions on observable quantities in order to obtain a predictivistic characterization of the multivariate and matrix-variate Student-t distributions as well as for the Student-t linear model. As a byproduct, a characterization for the Pearson type II distribution is provided.