The literature has covered the features and uses of the traditional univariate and bivariate logistic distributions in great detail. It is reasonable to wonder, though, if logistic marginals and conditionals could exhibit a similar behavior. A phenomenon that is comparable to both bivariate exponential and bivariate normal distributions. In this study, we will concentrate on bivariate distributions where one family of conditionals is marginal and the other family is of logistic type. Pseudo-logistic distributions are the name for such distributions. Research on conditionally specified models has revealed, however, that only in cases where the variables are independent will logistic marginals and both conditionals be of the logistic form occur. We talk about the features of distributional aspects and how they are built using the original. Both the original and the new conditioning regimes are used in two different ways. Possible generalizations are also considered. We also provide an example of a Pseudo-logistic model application.
翻译:文献已详细涵盖了传统单变量和双变量Logistic分布的特征及应用。然而,自然会提出一个问题:Logistic边缘分布和条件分布是否能够表现出类似的行为?这一现象与双变量指数分布和双变量正态分布具有可比性。在本研究中,我们聚焦于一类双变量分布,其中一组条件分布族是边缘的,而另一组条件分布族是Logistic类型的。这种分布被称为伪Logistic分布。然而,对条件指定模型的研究表明,仅当变量相互独立时,Logistic边缘分布和两个条件分布才会呈现Logistic形式。我们讨论了分布方面的特征,以及如何利用原始分布构建这些特征。原始和新的条件化方案以两种不同方式被使用。还考虑了可能的推广。我们同时提供了一个伪Logistic模型应用的实例。