Document Details

Document Type : Thesis 
Document Title :
Using Copula in Generalized Linear Model for Binary Data
استخدام الكوبيولا في النماذج الخطية المعممة للبيانات الثنائية
 
Subject : Faculty of Sciences 
Document Language : Arabic 
Abstract : Univariate probit and logit models are members of the family of generalized linear models (GLM), and analyze the relationship between regressors and binary response variable. The question of which model performs better is important, and the answer was achieved using a Monte Carlo simulation to compare both the univariate probit and logit models under different conditions. A bivariate probit model is frequently used in health economics, as its recursive form is practical when estimating the effect of binary endogenous (treatment) variable in a binary response model that is not suitable for univariate probit and logit models. Endogeneity can be controlled using a recursive bivariate probit model, which is completed by assuming that the structure errors of the model follow a standard bivariate Gaussian (Normal) distribution with correlation θ. The limitation of this model is its inability to deal with non-Gaussian dependencies between the treatment and response equations, and this is addressed by using a copula bivariate probit (CBP) model. This thesis will focus on the difference between probit and logit models in univariate case. In bivariate case, simulation has been applied to compare Gaussian, FGM and Frank copulas with symmetric dependency that are widely-used and well-known among copula researchers. 
Supervisor : Dr. Mervat Khalifa 
Thesis Type : Master Thesis 
Publishing Year : 1440 AH
2018 AD 
Co-Supervisor : Dr. Sulafah Binhimd 
Added Date : Monday, November 19, 2018 

Researchers

Researcher Name (Arabic)Researcher Name (English)Researcher TypeDr GradeEmail
عبير هادي السروجيAL-Sorouji, Abeer HadiResearcherMaster 

Files

File NameTypeDescription
 43820.pdf pdf 

Back To Researches Page