Model uncertainty quantification in Cox regression
Abstract
WeconsidercovariateselectionandtheensuingmodeluncertaintyaspectsinthecontextofCoxregression.Theperspectivewetakeisprobabilistic,andwehandleit within a Bayesian framework. One of the critical elements in variable/modelselection is choosing a suitable prior for model parameters. Here, we derive theso-called conventional prior approach and propose a comprehensive implemen-tation that results in an automatic procedure. Our simulation studies and realapplications show improvements over existing literature. For the sake of repro-ducibility but also for its intrinsic interest for practitioners, a web applicationrequiring minimum statistical knowledge implements the proposed approach.
Description
Ministerio de Ciencia e Innovación. Grant Number: Grant PID2019-104790GB-I00 funded by MCIN/AEI
Collections
- Artículos de Revista [1455]
