Online Program Home
  My Program

All Times EDT

Abstract Details

Activity Number: 45 - Nonparametric and Semiparametric Modeling for Complex Lifetime Data
Type: Contributed
Date/Time: Monday, August 3, 2020 : 10:00 AM to 2:00 PM
Sponsor: Lifetime Data Science Section
Abstract #312285
Title: Partially Linear Single-Index Generalized Mean Residual Life Models
Author(s): Peng Jin* and Mengling Liu
Companies: New York University and NYU Langone Medical Center
Keywords: Counting process; Estimating equation; Nonparametric regression; Spline; Survival analysis

Mean residual life (MRL) function is an important alternative to hazard function for characterizing time-to-event data. The MRL can be interpreted as remaining life expectancy of a subject who has survived to a time point. In biomedical research, the proportional MRL model has been primarily focused to study the association between risk factors and disease in multiplicative scale. When risk factors under investigation have complex correlation structures or are high-dimensional, the linearity assumption on the effects of covariates on log MRL function may be insufficient. The single-index (SI) model framework offers flexibility in capturing nonlinear covariate effects and reducing dimensionality. In this paper, we propose a family of partially linear single-index generalized MRL models, where the MRL function consists of both the nonparametric SI and the linear components. The spline technique is implemented to approximate the nonparametric SI function and estimate the parameters using an iterative algorithm. Asymptotic properties of the estimators are derived and the finite sample performance is evaluated through simulations. A real data application is presented for illustration.

Authors who are presenting talks have a * after their name.

Back to the full JSM 2020 program