EcoSta 2021: Start Registration
View Submission - EcoSta2021
A0658
Title: Hierarchical multi-parameter regression survival models Authors:  Fatima-Zahra Jaouimaa - University of Limerick (Ireland) [presenting]
Il Do Ha - Pukyong National University (Korea, South)
Kevin Burke - University of Limerick (Ireland)
Abstract: Standard survival models introduce covariates through a single (scale) parameter, and we refer to this standard practice as Single-Parameter Regression (SPR). In contrast, Multi-Parameter Regression (MPR) allows covariates to enter the model through multiple distributional parameters, i.e., scale and shape. Its flexibility has been highlighted in the context of survival data. We extend this to handle multivariate survival data by introducing random effects in both the scale and the shape regression components. We consider various possible dependence structures for these random effects (independent, shared, and correlated) and estimation proceeds using an $h$-likelihood approach. As the shape parameter may be viewed as a dispersion parameter for log-time, our proposal bears similarities to Double Hierarchical Generalized Linear Modelling (DHGLM). We investigate the performance of our estimation procedure using simulated data and also consider a real data example.