Comparación de procedimientos FDR para la selección de parámetros en Regresión Poisson
Comparison of FDR-based procedures to select parameters in Poisson Regression
Abstract (en)
The selection of significant variables in regression models is an important problem in applied statistics. Poisson Regression, useful when it is of interest to describe the number of occurrences of a particular event as a function of exploratory variables, has recently been used for modeling purposes in biology, epidemiology, genetics and engineering. Here, the Poisson Regression model as well as four procedures to select variables, all of them based on the False Discovery Rate (FDR), are described. In addition, these procedures are compared using a simulation study and some recommendations are given. As reference, the t-based and Bonferroni procedures were used. Finally, we model the number teenagers with children in the Department of Antioquia to illustrate these methods.
Abstract (es)
La selecci\'on de variables significativas en modelos de regresi\'on es un problema importante en el trabajo estad\'istico aplicado. El modelo de Regresi\'on Poisson, \'util para describir el n\'umero de ocurrencias de un evento particular como funci\'on de un conjunto de variables explicativas, ha sido recientemente empleado en biolog\'ia, epidemiolog\'ia, gen\'etica e ingenier\'ia. En este trabajo se describen el modelo de Regresi\'on Poisson y cuatro procedimientos para la selecciÛn de variables explicativas, todos basados en la tasa de falsos descubrimientos (FDR). Adicionalmente, estos procedimientos se comparan mediante un estudio de simulaci\'on y se dan algunas recomendaciones. Finalmente presentamos una aplicaci\'on donde se modela el n\'umero de madres menores de edad en el Departamento de Antioquia.
How to Cite
License
The authors maintain the rights to the articles and therefore they are free to share, copy, distribute, execute and publicly communicate the work under the following conditions:
Recognize the credits of the work in the manner specified by the author or licensor (but not in a way that suggests that, you have their support or that they support your use of their work).
Comunicaciones en Estadística is licensed under Creative Commons Atribución-NoComercial-CompartirIgual 4.0 Internacional (CC BY-NC-SA 4.0)
Universidad Santo Tomás preserves the patrimonial rights (copyright) of the published works, and favors and allows the reuse of them under the aforementioned license.
