Variable selection and estimation for the additive hazards model subject to left-truncation, right-censoring and measurement error in covariates

TL;DR

This paper develops a three-stage method for variable selection and estimation in high-dimensional additive hazards models, effectively addressing biases from left-truncation and measurement error in covariates.

Contribution

It introduces a novel three-stage procedure that simultaneously corrects measurement errors, selects variables, and estimates parameters in high-dimensional survival data with biases.

Findings

The proposed method improves variable selection accuracy.

Simulation studies demonstrate robustness against measurement error.

Method outperforms existing approaches in biased data scenarios.

Abstract

High-dimensional sparse modeling with censored survival data is of great practical importance, and several methods have been proposed for variable selection based on different models. However, the impact of biased sample caused by left-truncation and covariates measurement error to variable selection is not fully explored. In this paper, we mainly focus on the additive hazards model and analyze the high-dimensional survival data subject to left-truncation and measurement error in covariates. We develop the three-stage procedure to correct the error effect, select variables, and estimate the parameters of interest simultaneously. Numerical studies are reported to assess the performance of the proposed methods.