Premium
A quantile‐slicing approach for sufficient dimension reduction with censored responses
Author(s) -
Kim Hyungwoo,
Shin Seung Jun
Publication year - 2021
Publication title -
biometrical journal
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 1.108
H-Index - 63
eISSN - 1521-4036
pISSN - 0323-3847
DOI - 10.1002/bimj.201900250
Subject(s) - quantile , quantile regression , censoring (clinical trials) , slicing , sliced inverse regression , mathematics , dimension (graph theory) , statistics , dimensionality reduction , quantile function , sufficient dimension reduction , regression , computer science , econometrics , artificial intelligence , probability density function , cumulative distribution function , world wide web , pure mathematics
Sufficient dimension reduction (SDR) that effectively reduces the predictor dimension in regression has been popular in high‐dimensional data analysis. Under the presence of censoring, however, most existing SDR methods suffer. In this article, we propose a new algorithm to perform SDR with censored responses based on the quantile‐slicing scheme recently proposed by Kim et al. First, we estimate the conditional quantile function of the true survival time via the censored kernel quantile regression (Shin et al.) and then slice the data based on the estimated censored regression quantiles instead of the responses. Both simulated and real data analysis demonstrate promising performance of the proposed method.