Characterization of two competing risks models with Weibull risks

Competing risks are used to model life of systems with multiple failure modes where the observed failure is the minimum of the possible individual failures times. This research studies the effect ofignoring the risks by considering how well a single Weibull model approximates the competing risks model with Weibull risks. As a first approach to the problem this is done for a model with Weibull independent risks and for a specific model with Weibull dependent risks. Ignoring the risks may or not have a strong effect in the competing risks model and it would be of interest to characterize situations where ignoring the risks can cause serious biases in the competing risks model. The objective is to characterize the effect of the risks and their interdependence in the competing risks model, when the risks are Weibull distributed. Some recommendations are given on the correct use of these models.

• Autores: Yáñez Canal, Sergio
• Tipo de recurso: Doctoral thesis
• Fecha de publicación: 2015
• Institución: Universidad Nacional de Colombia
• Repositorio: Universidad Nacional de Colombia
• Idioma: spa
• OAI Identifier: oai:repositorio.unal.edu.co:unal/55265
• Acceso en línea: https://repositorio.unal.edu.co/handle/unal/55265; http://bdigital.unal.edu.co/50605/
• Palabra clave: 51 Matemáticas / Mathematics; 62 Ingeniería y operaciones afines / Engineering; Probability plots; Log-location-scale family; Weibull distribution; Multivariate Weibull with dependence; Lognormal distribution; Kullback-Leibler information; Ignoring-mode-of-failure; Maximum likelihood; Gráficos de probabilidad; Familia de log-localización-escala; Distribución Weibull; Distribución Weibull multivariada con dependencia; Distribución lognormal; Información de Kullback-Leibler; Ignorar modo de falla; Máxima verosimilitud
• Rights: openAccess
• License: Atribución-NoComercial 4.0 Internacional