New screening procedure for ultrahigh dimensional varying-coefficient model in longitudinal data analysis

Open Access
Author:
Chu, Wanghuan
Graduate Program:
Statistics
Degree:
Master of Science
Document Type:
Master Thesis
Date of Defense:
None
Committee Members:
  • Runze Li, Thesis Advisor
  • Matthew Logan Reimherr, Thesis Advisor
Keywords:
  • feature screening
  • varying-coefficient model
  • ultrahigh dimensionality
  • longitudinal data analysis
  • CAMP
Abstract:
This thesis is concerned with feature screening methods for varying-coefficient models in ultrahigh dimensional longitudinal setting. Motivated by an empirical analysis of the Childhood Asthma Management Project, CAMP, we introduce a new screening procedure for time-varying coefficient models with ultrahigh dimensional longitudinal predictor variables. The performance of the proposed procedure is investigated via Monte Carlo simulation. Numerical comparisons indicate that it can outperform existing ones substantially, resulting in significant improvements in explained variability and prediction error. Applying these methods to CAMP, we are able to find a number of potentially important genetic mutations related to lung function, several of which exhibit interesting nonlinear patterns around puberty.