Two alternative estimation procedures for the negative binomial cure rate model with a latent activation scheme
In this paper two alternative estimation procedures based on the EM algorithm are proposed forthe flexible negative binomial cure rate model with a latent activation scheme. The Weibull modelas well as the log-normal and gamma distributions are also considered for the time-to-event datafor the non-d...
| Autores: | , |
|---|---|
| Tipo de recurso: | artículo |
| Fecha de publicación: | 2016 |
| País: | España |
| Institución: | Universitat Politècnica de Catalunya (UPC) |
| Repositorio: | UPCommons. Portal del coneixement obert de la UPC |
| Idioma: | inglés |
| OAI Identifier: | oai:upcommons.upc.edu:2117/112736 |
| Acceso en línea: | https://hdl.handle.net/2117/112736 |
| Access Level: | acceso abierto |
| Palabra clave: | Competing risks EM algorithm latent activation scheme Classificació AMS::62 Statistics::62N Survival analysis and censored data Classificació AMS::62 Statistics::62P Applications Àrees temàtiques de la UPC::Matemàtiques i estadística::Estadística matemàtica |
| Sumario: | In this paper two alternative estimation procedures based on the EM algorithm are proposed forthe flexible negative binomial cure rate model with a latent activation scheme. The Weibull modelas well as the log-normal and gamma distributions are also considered for the time-to-event datafor the non-destroyed cells. Simulation studies show the satisfactory performance of the proposedmethodology. The impact of misspecifying the survival function on both components of the model(cured and susceptible) is also evaluated. The use of the new methodology is illustrated with areal data set related to a clinical trial on Phase III cutaneous melanoma patients. |
|---|