Quantile-based residual Matusita’s measure
摘要
In reliability and survival studies, assessing the similarity between the probability distributions of the residual lifetimes of two populations is of great importance. Matusita's overlapping measure is a widely used similarity measure based on the concept of distance between the probability distributions. In this paper, we consider the quantile-based Matusita’s measure for residual lifetimes. The quantile-based approach is particularly useful when one of the probability distributions being studied lacks an analytically tractable form, but its corresponding quantile function is explicitly defined. The properties of the proposed measure, including bounds and the effect of monotonic transformations, are discussed. A nonparametric estimator for the quantile-based residual Matusita's measure is also proposed. A simulation study is carried out to assess the performance of the proposed estimator.