Abstract
Median regression models provide a robust alternative to regression based on the mean. We propose a methodology for fitting a median regression model from data with both left and right censored observations, in which the left censoring variable is always observed. First we set up an adjusted least absolute deviation estimating function using the inverse censoring weighted approach, whose solution specifies the estimator. We derive the consistency and asymptotic normality of the proposed estimator and describe the inference procedure for the regression parameter. Finally, we check the finite sample performance of the proposed procedure through simulation.
Original language | English (US) |
---|---|
Pages (from-to) | 121-131 |
Number of pages | 11 |
Journal | Statistical Methodology |
Volume | 4 |
Issue number | 2 |
DOIs | |
State | Published - Apr 2007 |
Externally published | Yes |
All Science Journal Classification (ASJC) codes
- Statistics and Probability
Keywords
- Cox proportional hazards model
- Curse of dimensionality
- Doubly censored data
- Minimum dispersion statistic
- Missing information principle
- Volterra integral equation