The Multivariate Skewed Log-Birnbaum–Saunders Distribution and Its Associated Regression Model

In this article, a multivariate extension of the unit-sinh-normal (USHN) distribution is presented. The new distribution, which obtained from conditionally specified distributions methodology, absolutely continuous, and its marginal are univariate USHN. properties USHN studied in detail, statistical inference carried out classical approach using maximum likelihood method. suitable for modeling bounded data, especially (0,1)p region. addition, proposed extended to case regression model and, latter, Fisher information matrix derived. numerical results small simulation study two applications with real data sets allow us conclude that as well models, potentially useful analyze proportions, rates, or indices when them jointly considering different degrees correlation may exist variables interest.