ABSTRACT
Abstract This chapter introduces a novel copula-based methodology to analyze right censored bivariate survival data using flexible log-generalized extreme value (GEV) marginals. Copulas are quite popular in high-dimensional data modeling as they allow modeling and estimating the parameters of the marginal distributions separately from the dependence parameter of the copula. The Clayton copula structure has been used to represent the dependency between the log individual survival times for this chapter. We propose an empirical Bayes (EB) method for estimating the dependency parameter which has several benefits over the other existing methodologies. The EB methodology is implemented using an efficient importance sampling scheme.
