Weighted likelihood based inference for P(X < Y). Issue 10 (26th November 2017)

• Record Type: Journal Article
• Authors: Greco, Luca
• Is Part Of: Communications in statistics. Volume 46:Issue 10(2017)
• Page Start: 7777
• Page End: 7789
• Publication Date: 2017-11-26
• Subjects: Outliers -- Robustness -- Stress–strength model -- Weighted likelihood; Primary 62F35 -- Secondary 62F03; Mathematical statistics -- Periodicals; Mathematical statistics -- Data processing -- Periodicals; Digital computer simulation -- Periodicals; 519.5
• DOI: 10.1080/03610918.2016.1252396
• Languages: English
• ISSNs: 0361-0918

Abstract

ABSTRACT: This contribution deals with the statistical problem of evaluating the stress–strength reliability parameter R = P ( X < Y ), when both stress and strength data are prone to contamination. Standard likelihood inference can be badly affected by mild data inadequacies, that often occur in the form of several outliers. Then, robust tools are recommended. Here, inference relies on the weighted likelihood methodology. This approach has the advantage to lead to robust estimators, tests, and confidence intervals that share the main asymptotic properties of their classical counterparts. The accuracy of the proposed methodology is illustrated both by numerical studies and real-data applications.