Publication Details
Publisher: Central Asian Studies
Issue: Vol 8, No 1 (2026)
ISSN: 2660-4159

Abstract

The research aims at filling a critical gap in the field of applied statistics, which is to come up with a measure of uncertainty in censored data, especially in survival analysis. The main aim of the research is to develop a non-parametric approach to estimating the “Residual Extropy”. It is a new measure that acts as a complement to entropy, with a specific focus on the uncertainty in the tail part of the distribution, but considering censored data. Kaplan-Meier estimation approach was used to estimate the survival function, which helped to effectively deal with censored data. Following this, the kernel estimator was used to smooth the survival function and to estimate the probability density function. It is an important step that sets the stage for estimation of the extropy. The findings revealed a significant connection between the survival function, hazard (mortality) density, and residual uncertainty. We found that the level of uncertainty reaches its highest during the periods with the highest mortality rates (the peak of the mortality density), This proves that extropy is a sensitive and effective measure for capturing hazard dynamics. It was discovered that residual extropy estimation is a valuable technique when used in the non-parametric setting, giving medical practitioners and others in the field an ability to make better decisions through a more profound comprehension of uncertainty in survival data.

Keywords
Residual Extropy Nonparametric Estimation Censored Data Survival Analysis Kaplan–Meier Estimator Kernel Estimator