Inferencia estatística con datos censurados
Loading...
Identifiers
Publication date
Authors
Advisors
Editors
Journal Title
Journal ISSN
Volume Title
Publisher
Abstract
Neste traballo estúdanse algunhas das ferramentas máis habituais na análise de datos con censura, enmarcadas dentro da rama da estatística denominada análise de supervivencia. Para comezar, introducíronse unha serie de conceptos, definicións e funcións básicas recurrentes ao longo do traballo. A continuación, explicouse o método de estimación de modelos paramétricos mediante máxima verosimilitude con datos censurados. Ademais, facendo uso de R fíxose un estudo de simulación para analizar as propiedades dos estimadores paramétricos, dos intervalos de confianza asintóticos e por perfil de verosimilitude. Máis adiante, presentáronse os estimadores non paramétricos da distribución, primeiro o que se obtén coas táboas de vida e despois mediante o estimador de Kaplan-Meier. Para rematar, mostrouse unha proba gráfica de bondade de axuste dos modelos paramétricos baseada no estimador de Kaplan-Meier. Ilustráronse estes métodos cos datos pharmacoSmoking do paquete asaur
In this project some of the most common tools for the analysis of censored data are studied, framed within the branch of statistics known as survival analysis. To begin, we introduced some basic concepts, definitions and functions that appear throughout the project. Next, the estimation method based on maximum likelihood with censored data for parametric models was explained. Besides, a simulation study was made using R, analyzing properties of parametric estimates, asymptotic confidence intervals and by profile likelihood. Later, non-parametric estimates of the distribution are presented, firstly using life tables and then using the Kaplan-Meier estimate. To conclude, a graphic method to analyze the goodnes-of-fit of parametric models based on the Kaplan-Meier estimate was shown. All these methods were exemplified with pharmacoSmoking data from asaur package
In this project some of the most common tools for the analysis of censored data are studied, framed within the branch of statistics known as survival analysis. To begin, we introduced some basic concepts, definitions and functions that appear throughout the project. Next, the estimation method based on maximum likelihood with censored data for parametric models was explained. Besides, a simulation study was made using R, analyzing properties of parametric estimates, asymptotic confidence intervals and by profile likelihood. Later, non-parametric estimates of the distribution are presented, firstly using life tables and then using the Kaplan-Meier estimate. To conclude, a graphic method to analyze the goodnes-of-fit of parametric models based on the Kaplan-Meier estimate was shown. All these methods were exemplified with pharmacoSmoking data from asaur package
Description
Traballo Fin de Grao en Matemáticas. Curso 2021-2022
Keywords
Bibliographic citation
Relation
Has part
Has version
Is based on
Is part of
Is referenced by
Is version of
Requires
Sponsors
Rights
Atribución-NoComercial-CompartirIgual 4.0 Internacional



