Some alternative bivariate Kumaraswamy models. Issue 18 (17th September 2017)

• Record Type: Journal Article
• Authors: Arnold, Barry C.; Ghosh, Indranil
• Is Part Of: Communications in statistics. Volume 46:Issue 18(2017)
• Page Start: 9335
• Page End: 9354
• Publication Date: 2017-09-17
• Subjects: 60E, 62F; Mathematical statistics -- Periodicals; Mathematics; Statistics; 519.2
• DOI: 10.1080/03610926.2016.1208244
• Languages: English
• ISSNs: 0361-0926

Abstract

ABSTRACT: In this article we discuss various strategies for constructing bivariate Kumaraswamy distributions. As alternatives to the Nadarajah et al. (2011) bivariate model, four different models are introduced utilizing a conditional specification approach, a conditional survival function approach, and an Arnold–Ng bivariate beta distribution construction approach. Distributional properties for such bivariate distributions are investigated. Parameter estimation strategies for the models are discussed, as are the consequences of fitting two of the models to a particular data set involving the proportion of foggy days at two different airports in Colombia.