Extensões da distribuição de Lindley para análise de dados de sobrevivência

• Ano de defesa: 2014
• Autor(a) principal: Espirito Santo, Ana Paula Jorge do [UNESP]
• Tipo de documento: Dissertação
• Tipo de acesso: Acesso aberto
• Idioma: por
• Instituição de defesa: Universidade Estadual Paulista (Unesp)
• Palavras-chave em Português: Computação - Matematica; Análise de sobrevivência (Biometria); Riscos competitivos; Sistemas de parâmetros distribuídos; Estatistica - Processamento de dados; Computer science - Mathematics
• Link de acesso: http://hdl.handle.net/11449/113833

Survival Analysis is an area of Statistics that has developed over the last years and many probability distributions have been proposed . In this work, three extensions of Lindley distribution of one parameter obtained by the composition of the distribution process considering the latent structures of activation , minimum and maximum were presented . It is noteworthy that in a context of competing risks or additional risks where it is assumed that the number of causes of failures M follows a Geometric distribution , the distribution obtained by the composition process is a particular case of Marshall-Olkin Extended distribution . The new distributions proposed in this work are: Marshall-Olkin Extended Lindley distribution (Lindley-Geometric distribution), Zero Truncated Lindley-Poisson distribution and Size- Biased Lindley-Poisson distribution. For characterization of distributions some mathematical properties and six estimation methods were presented.