Goodness-of-fit for Biparametric k-Modified models

• Grant number: 18/04053-5
• Support Opportunities: Scholarships in Brazil - Scientific Initiation
• Start date: May 01, 2018
• End date: April 30, 2019
• Field of knowledge: Physical Sciences and Mathematics - Probability and Statistics - Applied Probability and Statistics
• Principal Investigator: Katiane Silva Conceição
• Grantee: Maria Pinheiro Garcia Blanco
• Host Institution: Instituto de Ciências Matemáticas e de Computação (ICMC). Universidade de São Paulo (USP). São Carlos , SP, Brazil

Abstract

In this project, we will consider some criteria that evaluate the goodness-of-fit of the k-modified regression models. The k-modification consists of the inclusion of a parameter in the probability mass function of the traditional discrete distributions, capable of modeling the inflation or deflation of the observation k in the data set. In this sense, the modification becomes essential when, in many practical situations, a particular observation k of the data set occurs with a frequency greater or less than expected when considering a given traditional discrete distribution. In the context of the regression models, we want to explain the k-modified count data set using explanatory variables for the mean and for the parameter responsible for the modification of the observation k. In this work, we will consider the following k-modified models: Poisson, Binomial and Geometric. (AU)