Hazard rate estimation when the measurement error has a normal or logistic distribution

TL;DR

This paper studies how to estimate hazard rates when data has normal or logistic measurement errors, improving reliability analysis.

Contribution

The paper introduces a method for hazard rate estimation with measurement errors from normal or logistic distributions.

Findings

Local time polynomial estimators were used to estimate the density function.

The risk rate function was estimated under 15% and 30% contamination levels.

Numerical analysis demonstrated the effectiveness of the proposed method.

Abstract

Statistical data analysis available in most scientific fields is often recorded with measurement error. The modeling of these statistical data by ignoring the measurement errors, leads to estimators of the parameters of the distributions, whose use does not achieve sufficient accuracy in the goodness of fit. In reliability criteria, one of the important issues is hazard rate function. It prompted us to investigate the hazard rate criterion in the presence of measurement error generated from the normal or logistic distribution. Now, while providing the estimator for the density function using local time polynomial estimator methods, the risk rate function is estimated according to the contamination degree of 15 or 30%. Finally, we present the numerical analysis.

Figures