Improved Bayes estimators and prediction for the Wilson-Hilferty distribution

Autor(es)

Ramos, Pedro L. ^[1] ; Almeida, Marco P. ^{[1, 2]} ; Tomazella, Vera L. D. ^[2] ; Louzada, Francisco ^[1]

Afiliação do(s) autor(es)

^[1] Univ Sao Paulo, Dept Matemat Aplicada & Estat, Ave TrabalhadorSao Carlense, 400, BR-13566590 Sao Carlos, SP - Brazil

^[2] Univ Fed Sao Carlos, Dept Estat, Rodovia Washington Luis, Km 235, BR-13565905 Sao Carlos, SP - Brazil

• Tipo de documento: Artigo Científico
• Fonte: Anais da Academia Brasileira de Ciências; v. 91, n. 3 2019.
• Citações Web of Science: 0

Resumo

Abstract: In this paper, we revisit the Wilson-Hilferty distribution and presented its mathematical properties such as the r-th moments and reliability properties. The parameters estimators are discussed using objective reference Bayesian analysis for both complete and censored data where the resulting marginal posterior intervals have accurate frequentist coverage. A simulation study is presented to compare the performance of the proposed estimators with the frequentist approach where it is observed a clear advantage for the Bayesian method. Finally, the proposed methodology is illustrated on three real datasets. (AU)

• Processo FAPESP: 17/25971-0 - Inferência estatística de sistemas complexos
• Beneficiário: Pedro Luiz Ramos
• Modalidade de apoio: Bolsas no Brasil - Pós-Doutorado