Executive Summary : | Regression analysis is one of the widely used statistical techniques and plays a vital role in studying the dependence between response and predictor variables. Classical regression models have limited real situation applications as these models have various assumptions on the dependent variable. Copula is a powerful tool for modeling dependence among random variables. It has applications in finance, medicine, engineering, agriculture and geophysics. Copula-based regression models are more flexible and overcome several drawbacks associated with classical regression and generalized linear. The main focus of this research proposal is to develop some new, improved estimation procedures for copula-based regression models. It also aims to study the performances of proposed estimators using numerical experiments under specified and misspecified scenarios. Researchers intend to discuss various statistical properties like unbiasedness, consistency, and asymptotic properties of the estimators with real applications. |