The purpose of this project is to present a study of inference and Diagnostics analysis in bivariate regression model Birnbaum-Saunders (BS) based on the scale mixture on normal (SMN) distributions.This model represents a robust extension to the context of the bivariate regression model proposed by Rick and Nedelman (1991). This line of study is a supplement to works of Kundu et al. (2010), Kundu et al. (2013), and Vilca et al. (2013). We will be following the same ways considered in the univariate log-linear regressiom BS, both from view point of the estimation as well as the diagnostics analysis (Cook, 1986; Zhu and Lee, 2001).
News published in Agência FAPESP Newsletter about the scholarship: