Figure 2

Model description with all the parameters involved for a single cluster. The complete hierarchical model with the parametrization for a single cluster model. Depicted in blue are the hyperparameters for the respective distributions, like (r, s) for the Gamma prior on ρ. The observed variables x denoting the covariates and t denoting time are shown in green. The part of the figure centered around t forms the core which defines the generalized linear model with a Normal random link between η and the covariates and coefficients and priors for the Weibull distribution. The block on the right defines the hierarchy related to the sparse regression on the covariates via the hierarchical representation of the Normal-Gamma prior on the regression coefficients β. Furthermore, the left block defines the variables for describing the distribution of the covariate space.