Statistical Inference for Treatment Effects in High-Dimensional AFT Model

• Activity Number: 29
• Type: Contributed
• Date/Time: Sunday, August 3, 2014 : 2:00 PM to 3:50 PM
• Sponsor: Biometrics Section
• Author(s): Hao Chai*+ and Jian Huang
• Companies: University of Iowa and University of Iowa
• Keywords: Confidence interval ; Asymptotic normality ; LASSO ; Accelerated failure time model ; Sign consistency

Abstract

We consider the problem of estimating a low-dimensional treatment effect parameter in the presence of a large number of potentially confounding variables in the accelerated failure time model. Our approach first estimates the effects of the high-dimensional confounding variables and selects a parsimonious model using penalized methods. Then a plug-in method is used to estimate the treatment effects. The asymptotic normality of the treatment effects estimator is derived under certain regularity conditions, and it holds despite the inconsistency in the selection of the high-dimensional nuisance parameters. The performance of the proposed method is demonstrated using simulations and real data examples.