In this paper a new family of statistics based on -divergence for testing goodness-of-fit under composite null hypotheses are considered. The asymptotic distribution of this test is obtained when the unspecified parameters are estimated by maximum likelihood as well as minimum -divergence.
In this paper a bayesian criterion for comparing different experiments based on the maximization of the f*-Divergence is proposed and studied. After a general setting of the criterion, we prove that this criterion verifies the main properties that a criterion for comparing experiments must satisfy.
A statistic using the concept of order - weighted information energy introduced by Tuteja et al. (1992) is considered and its asymptotic distribution in a stratified random sampling is obtained. Some special cases are also discussed.
En esta comunicación se establece una medida de la entropía contenida en un proceso puntual mediante el concepto de entropía de orden α y tipo β introducida por Sharma and Mittal (1975); quedando, de este modo, generalizada la entropía de McFadden. Una vez que se estudian las propiedades relativas a la tasa de cambio de la Entropía, se demuestra que el proceso de Poisson es el de Entropía máxima dentro de la clase de los procesos puntuales estacionarios.
In this paper we consider an exploratory canonical analysis approach for multinomial population based on the -divergence measure. We define the restricted minimum -divergence estimator, which is seen to be a generalization of the restricted maximum likelihood estimator. This estimator is then used in -divergence goodness-of-fit statistics which is the basis of two new families of statistics for solving the problem of selecting the number of significant correlations as well as the appropriateness...
Page 1