Bazı normal olmayan dağılımlar için istatistiksel sonuç çıkarımı
Özet
In this thesis, parameter estimattors of Kumaraswamy Normal (KwNormal) and Kumaraswamy Weibull (KwWeibull) distributions are considered. Maximum Likelihood (ML), Modified Maximum Likelihood (MML), Least Squares (LS), Weighted Least Squares (WLS), Cramer- The von Mises (CM) and Anderson Darling (AD) estimators of the unknown parameters of KwNormal are obtained. ML, LS, WLS, CM and AD estimators of the unknown parameters of KwWeibull are obtained. A Monte Carlo simulation study is conducted to compare the efficiencies of these estimators. The results of the Monte-Carlo simulation study demonstrate that ML, MML and AD estimators of the parameters of the KwNormal distribution are more efficient than the corresponding LS, WLS and CM estimators. ML and AD estimators of the parameters of the KwWeibull distribution are more efficient than the corresponding LS, WLS and CM estimators. At the end of the study, KwNormal and KwWeibull distributions are used to model the real life data set taken from the literature.