Resumen Las distribuciones de probabilidad asimétricas han sido ampliamente estudiadas por diversos autores en las últimas décadas. Se ha tenido especial interés en familias de distribuciones flexibles con la capacidad de tener en cuenta grados de asimetría y curtosis mayores que las distribuciones el ampliamente conocidas en teoría estadística. Si bien la mayoría de las nuevas distribuciones se ajustan a datos unimodales y unas pocas a datos bimodales, en las propuestas bimodales se han encontrado problemas de singularidad en las matrices de información. Por lo tanto, en este artículo se desarrollan extensiones de la familia de distribuciones alfa-potencia, que tienen matriz de información no singular. Las nuevas propuestas se basan en las distribuciones bimodal-normal y bimodal elíptica sesgada-normal. Estas nuevas extensiones permiten modelar datos bimodales asimétricos, que se encuentran comúnmente en varias áreas de interés científico. También se estudian en detalle las propiedades de estas nuevas distribuciones de probabilidad, y se realiza el proceso de inferencia estadística para estimar los parámetros de los modelos propuestos. La convergencia estocástica para el vector estimador de máxima verosimilitud (EMV) se puede encontrar debido a la no singularidad de la matriz de información esperada en el soporte correspondiente. También introdujimos extensiones de los modelos asimétrico bimodal normal y bimodal elíptico sesgado-normal para las situaciones en las que los datos presentan censura. También se presenta un pequeño estudio de simulación para evaluar las propiedades del EMV y, finalmente, se presentan dos aplicaciones a conjuntos de datos reales con fines ilustrativos.

Abstract Asymmetric probability distributions have been widely studied by various authors in recent decades. Special interest has been had families of flexible distributions with the capability to have into account degree of skewness and kurtosis greater than the cl 1 distributions widely known in statistical theory. While, most of the new distributions fit unimodal data, and a few fit bimodal data, in the bimodal proposals, singularity problems have been found in the information matrices. Therefore, in this paper, extensions of the alpha-power family of distributions are developed, which have non-singular information matrix. The new proposals are based on the bimodal-normal and bimodal elliptical skew-normal distributions. These new extensions allow modeling asymmetric bimodal data, which are commonly found in several areas of scientific interest. The properties of these new distributions of probability are also studied in detail, and the statistical inference process is carried out to estimate the parameters of the proposed models. The stochastic convergence for the maximum likelihood estimator (MLE) vector can be found due to the non-singularity of the expected information matrix in the corresponding support. We also introduced extensions of the asymmetric bimodal normal and bimodal elliptical skew-normal models for the situations in which the data present censorship. A small simulation study to evaluate the properties of the MLE is also presented and, finally, two applications to real data set are presented for illustrative purposes.

Abstract This paper introduce two new families of distributions that allow fitting unimodal, bimodal or trimodal data sets. Statistical properties such as distribution function, moments, moment generating function and stochastic representation of these new families are studied in details. The problem of estimating parameters is addressed by considering the maximum likelihood method and Fisher information matrices are derived. A small Monte Carlo simulation study is conducted to examine the performance of the obtained estimators. The methodology developed is illustrated with three real data applications.

Abstract In this paper, we propose the power Student-t regression model for censored (limited) observations which extends the Student-t censored regression model. This extension is based on the asymmetric and heavy-tailed power Student-t distribution. The score functions and expected information matrix are given as well as the process for estimating the parameters in the model is discussed by using the likelihood approach. Two simulation studies are conducted to evaluate parameter recovery and properties of the model and finally, two applications to a real data set are reported to demonstrate the usefulness of this new methodology.