New continuous distributions applied to lifetime data and survival analysis

• Ano de defesa: 2016
• Autor(a) principal: DIAS, Cícero Rafael Barros
• Tipo de documento: Tese
• Tipo de acesso: Acesso aberto
• Idioma: por
• Instituição de defesa: Universidade Federal de Pernambuco; UFPE; Brasil; Programa de Pos Graduacao em Estatistica
• Palavras-chave em Português: Estatística.; Estimadores de máxima; Máxima verossimilhança
• Link de acesso: https://repositorio.ufpe.br/handle/123456789/17307

Statistical analysis of lifetime data is an important topic in engineering, biomedical, social sciences and others areas. There is a clear need for extended forms of the classical distributions to obtain more flexible distributions with better fits. In this work, we study and propose new distributions and new classes of continuous distributions. We present the work in three independentes parts. In the first one, we study with some details a lifetime model of the beta generated class proposed by Eugene; Lee; Famoye (2002). The new distribution is called the beta Nadarajah-Haghighi distribution, which can be used to model survival data. Its failure rate function is quite flexible and takes several forms depending on its parameters. The proposed model includes as special models several important distributions discussed in the literature, such as the exponential, generalized exponential (GUPTA; KUNDU, 1999), extended exponential (NADARAJAH; HAGHIGHI, 2011) and exponential-type (LEMONTE, 2013) distributions. We provide a comprehensive mathematical treatment of the new distribution and obtain explicit expressions for the moments, generating and quantile functions, incomplete moments, order statistics and entropies. The method of maximum likelihood is used for estimating the model parameters and the observed information matrix is derived. We fit the proposed model to a real data set to prove empirically its flexibility and potentiality. In the second part, we study general mathematical properties of a new generator of continuous distributions with three extra shape parameters called the exponentiated Marshal-Olkin family. We present some special models of the new class and some of its mathematical properties including moments and generating function. The method of maximum likelihood is used for estimating the model parameters. We illustrate the usefulness of the new distributions by means of two applications to real data sets. In the third part, we propose another new class of distributions based on the distribution introduced by Nadarajah and Haghighi (2011). We study some mathematical properties of this new class called Nadarajah-Haghighi-G (NH-G) family of distributions. Some special models are presented and we obtain explicit expressions for the quantile function, ordinary and incomplete moments, generating function and order statistics. The estimation of the model parameters is explored by maximum likelihood and we illustrate the flexibility of the new family with two applications to real data.