Search

In the analysis of financial market data, random variations in price movements can share non-trivial statistical properties as observed for the distributions of returns, with possibility of absence of autocorrelations in asset returns, volatility in blocks and presence of asymmetry between rises and falls of stock values. Asset price fluctuations are usually investigated using time series models to obtain inferences about interest rates and future forecasts. In this study we consider the use of existing probability distributions to model run lengths and absolute historical price run returns as an alternative to the use of usual time series models with applications for time series obtained from the NYSE stock exchange for three private banks located in the Brazil in the period from July 19, 2013, to July 19, 2018. We assume discrete Weibull distributions as an alternative to the exponential law commonly used in this type of analysis. Under this modeling approach it is possible to obtain information about the structure of the market such as the probability of the stock market rising and falling daily and the magnitude of the returns. [ABSTRACT FROM AUTHOR]

In many practical situations, clinical diagnostic procedures include two or more biological traits whose outcomes are expressed on a continuous scale and are then dichotomized using a cut point. As measurements are performed on the same individual there is a likely correlation between the continuous underlying traits that can go unnoticed when the parameter estimation is done with the resulting binary variables. In this paper, we compare the performance of two different indexes developed to evaluate the dependence between diagnostic clinical tests that assume binary structure in the results with the performance of the binary covariance and two copula dependence parameters.Muchos procedimientos de diagnóstico clínico médico exigen la evaluación de dos o mas rasgos biológicos que se ven alterados ante la presencia de fenómenos de enfermedad o infección, los cuales se expresan en una escala continúa de medición con posterior dicotomización usando de un valor límite o punto de corte. Dado que las mediciones son realizadas en el mismo individuo, los resultados probablemente presenten dependencia de algún tipo, lo cual puede ser ignorado en la etapa de análisis de datos dada la presentación binaria de los datos. En este estudio comparamos el comportamiento de dos parámetros de dependencia presentes en funciones de cópula con el de la covarianza binaria y dos índices creados para medir dependencia entre pruebas diagnósticas de respuesta dicótoma.

In this paper, we show different parameters estimation forms for multiple linear regression model. We used clinical data, where the interest was to verify the relationship among the mechanical assay maximum stress with femoral mass, femoral diameter and group of ovariectomized Wistar rats. We used three inference methods: Classic inference, based on the least square method; bayesian inference, based on the Bayes theorem; and bootstrap inference, based on resampling processes.En este trabajo se muestran diferentes formas de estimación de parámetros para el modelo de regresión lineal múltiple. Para estimar los parámetros del modelo se utilizaron los datos de un ensayo clínico donde el interés era verificar la relación entre la tensión mecánica máxima del fémur con la masa y el diámetro femoral en un grupo experimental de ratas ovariectomizadas de la raza Rattus norvegicus albinos, variedad Wistar. Para estimar los parámetros del modelo se compararon tres métodos: la metodología clásica basada en el método de mínimos cuadrados, la metodología bayesiana, basada en el teorema de Bayes y la inferencia bootstrap basada en los procesos de remuestreo.Nesse artigo, tem-se o interesse em avaliar diferentes estratégias de estimação de parâmetros para um modelo de regressão linear múltipla. Para a estimação dos parâmetros do modelo foram utilizados dados de um ensaio clínico em que o interesse foi verificar se o ensaio mecânico da propriedade de força máxima (EM-FM) está associada com a massa femoral, com o diâmetro femoral e com o grupo experimental de ratas ovariectomizadas da raça Rattus norvegicus albinus, variedade Wistar. Para a estimação dos parâmetros do modelo serão comparadas três metodologias: a metodologia clássica, baseada no método dos mínimos quadrados; a metodologia Bayesiana, baseada no teorema de Bayes; e o método Bootstrap, baseado em processos de reamostragem.

En este artículo se presenta una breve descripción de modelos GARCH multivariados y se realizan inferencias de la volatilidad de series de tiempo usando un enfoque Bayesiano, utilizando algoritmos de simulación de Monte Carlo (MCMC). Como una aplicación para ilustrar la metodología propuesta, se analizan los log-retornos de IBOVESPA y Dow Jones Industrial en una base semanal de 04/27/1993 para 11/03/2008.

In this study, the components of extra-Poisson variability are estimated assuming random effect models under a Bayesian approach. A standard existing methodology to estimate extra-Poisson variability assumes a negative binomial distribution. The obtained results show that using the proposed random effect model it is possible to get more accurate estimates for the extra-Poisson variability components when compared to the use of a negative binomial distribution where it is possible to estimate only one component of extra-Poisson variability. Some illustrative examples are introduced considering real data sets. [ABSTRACT FROM AUTHOR]