Banca de DEFESA: JAINE DE MOURA CARVALHO

Uma banca de DEFESA de MESTRADO foi cadastrada pelo programa.
STUDENT : JAINE DE MOURA CARVALHO
DATE: 28/02/2023
TIME: 14:00
LOCAL: http://meet.google.com/xxa-zgmu-zob
TITLE:

Slashed Lomax Distribution: Goodness-of-fit measures through the Mellin transform

KEY WORDS:

Slashed Lomax Distribution, Mellin Transform, Log-Cumulants, Goodness-of-fit measures, Hotelling's T^2 statistic.

PAGES: 42
BIG AREA: Ciências Agrárias
AREA: Agronomia
SUMMARY:

Recently, various probability distributions have been proposed to achieve satisfactory results, specifically models with increased flexibility that can model data on the duration of components or the lifetime of individuals. Among these, the Slashed class models, particularly the Slashed Lomax distribution, have gained attention. This asymmetric model is defined for positive real values, and is notable for its stochastic representation and ability to fit heavy-tailed data sets. Despite the increasing number of new probabilistic models that cater to specific samples, there have been few statistical tools introduced to evaluate their goodness of fit. To address this deficit, we employed the methodology outlined in Nicolas (2002) and utilized second-type statistics (log-cumulative) derived from the Mellin Transform (TM) to provide new measures of goodness of fit for the Slashed Lomax distribution. These measures consider both qualitative and quantitative aspects. We derived the TM expression for the Slashed Lomax distribution, calculated the log-cumulants (LCs) and created the LC diagram (k ̃_3,k ̃_2). Then, we proposed a test statistic using a combination of Hotelling's T^2 statistic and the multivariate Delta method to test hypotheses about the LCs. Finally, we applied the proposed methodology to two real databases in the context of survival analysis to show its effectiveness in evaluating the fit criteria. We conducted bootstrap experiments to assess the power of the proposed test and to evaluate the performance of the estimators using the log-cumulative method (MLC), method of moments (MM), and maximum likelihood method (ML). The results revealed that the adjustment tools performed well, and that the MLC proved to be an effective estimation criterion.

COMMITTEE MEMBERS:
Externo à Instituição - JHONNATA BEZERRA DE CARVALHO - UFAL
Presidente - FRANK SINATRA GOMES DA SILVA
Interno - JOSIMAR MENDES DE VASCONCELOS