Regression models based in the generalized gamma distribution

• Grant number: 12/19410-1
• Support Opportunities: Scholarships abroad - Research
• Start date: January 02, 2013
• End date: July 01, 2013
• Field of knowledge: Physical Sciences and Mathematics - Probability and Statistics - Statistics
• Principal Investigator: Gladys Dorotea Cacsire Barriga
• Grantee: Gladys Dorotea Cacsire Barriga
• Host Investigator: Dipak Kumar Dey
• Host Institution: Administração Geral. Universidade Estadual Paulista (UNESP). Campus de Bauru. Bauru , SP, Brazil
• Institution abroad: University of Connecticut (UCONN), United States

Abstract

In this project we propose two regression models, the first is the regression model log-generalized gamma (LGG) with measurement error and the second model Regression zeros inflated power series. In this second model we consider as link function, to proportion of zeros the cumulative distribution function of the distribution LGG. Moreover, the proposed models we intend to develop inferential procedures and diagnostic analysis, from classical and Bayesian perspective. (AU)