On estimation of extropy for non-negative data with application on uniformity testing

Title

On estimation of extropy for non-negative data with application on uniformity testing

Creator

Khudania, Harshita; Rajesh, Richu; Raju, David Chris

Description

{Poisson weights-based density estimator is used to estimate the extropy function to the non-negative data}. The traditional class of nonparametric extropy estimators, typically constructed using kernel density estimators with symmetric kernels, is not well suited for non-negative data. To address this limitation, we propose two Poisson-weights-based density estimators that are naturally adapted to the non-negative domain. The asymptotic properties of the proposed estimators are rigorously established, providing theoretical support for their use. A comprehensive simulation study demonstrates that both estimators outperform their conventional kernel-based counterparts in terms of bias and mean squared error. Furthermore, we introduce uniformity tests based on extropy and obtain their critical values through simulation. The practical utility of the proposed methods is illustrated through analyses of real data sets. 2025 Informa UK Limited, trading as Taylor & Francis Group.

Source

Journal of Statistical Computation and Simulation;Volume;96;Issue;7;pp.1709-1727

Date

01-01-2026

Publisher

Taylor and Francis Ltd.

Subject

Extropy; Kernel estimator; mean squared error; Poisson weights; uniform testing

Coverage

Khudania H., Department of Statistics and Data Science, Christ (Deemed to be University), Karnataka, Bengaluru, India; Rajesh R., Department of Statistics, Government Victoria College, Kerala, Palakkad, India; Raju D.C., Department of Statistics, Nirmala College Muvattupuzha, Kerala, Ernakulam, India

Rights

Restricted Access; Hardcopy may be available in the library

Relation

ISSN: 949655;

Format

online

Language

English

Type

Article

Identifier

https://doi.org/10.1080/00949655.2025.2605498

https://www.scopus.com/pages/publications/105025469866?origin=resultslist