Biostatistics and Bioinformatics Monthly Seminar with Xianming Tan, PhD
“Nonparametric Convolution Regression Models and Its Applications in Clinical Studies”
Xianming Tan, PhD
Research Professor
Department of Biostatistics and Lineberger Comprehensive Cancer Center
The University of North Carolina at Chapel Hill
Meeting ID: 834 2878 7778
Passcode: Sy6FTVFi
Abstract:
We introduce an extended model that harnesses the power of convolution operations to represent time-varying treatment and carry-over effects in a crossover study design or other designs like N-of-1 design. Unlike the traditional model, the proposed approach unifies the treatment and carry-over effects through time-varying response functions, one for each treatment. The model is not only flexible enough to accommodate a variety of treatment plans, including multiple administrations at different doses, but also allows for the inclusion of more treatment periods. The advantages of this approach are accentuated by its ability to be generalized, to avoid prior assumptions about the carry-over effect and to maintain consistent estimation and hypothesis testing procedures. In this paper, we explore the details of hypothesis testing under this extended model, focusing on the comparison of two response functions within specified intervals. The goal of this work is to improve the modeling of carry-over effects, thereby strengthening the applicability of the model to a variety of experimental settings.
This group meets monthly from Noon – 1 p.m.