![]() | Only 14 pages are availabe for public view |
Abstract Regression models generally assume one parametric form for the whole domain of interest however this might not be the case when a change-point exists. Consequently, the change-point’s existence needs to be investigated and estimated if it does occur. Plenty of the models studying the change-point problem assumed normal responses. However, the normality assumption would be violated in many cases such as heavy tailed data and in the presence of outliers. In such cases, quantile regression model is a suitable alternative where it is known to be distribution free and robust to outliers. CUSUM test is one of the oldest and most known change-point tests. It has been used to detect the existence of the threshold effect to the quantile regression model but using cross-sectional data.This thesis presents an extension of the CUSUM test for investigating the existence of a change-point in a quantile regression model to the longitudinal data setting in the complete case. Simulation studies are used to assess the performance of the proposed test. Finally, the proposed test is used to detect any possible change-point in COVID- 19 dataset |