Quantile Mixture Cure Model and Its Estimation Methods
This presentation will feature Dr. Yingwei (Paul) Peng, Professor in the Department of Public Health Sciences and Department of Mathematics and Statistics at Queen’s University, Canada.
About
The Biostatistics seminar series invites researchers from across the nation to discuss methodological research and its implications for a variety of health issues.
Abstract
Quantile regression for survival data received a great deal of recent attention and it has been considered in modeling the failure time distribution of uncured subjects in the mixture cure model. In this talk, I will review existing methods for estimating the cure model and propose a novel estimating equation approach. The approach does not rely on the global linearity assumption, and it allows partially functional covariate effects in the quantile regression part. The estimation method also enjoys a double robustness property in the sense that the latency estimation does not depend on the incidence estimation. The properties of the proposed method are demonstrated in a simulation study. The application of the proposed method to a lung cancer dataset reveals new findings in comparison with existing methods.
Contact
Andy Ni