Optimal discrimination design for copula models

TL;DR

This paper introduces a new discrimination design approach using $D_s$-optimality to effectively select the correct dependence structure in copula models, enhancing model discrimination capabilities.

Contribution

It extends optimal design theory for copula models by applying $D_s$-optimality for copula selection, providing a novel method for dependence structure discrimination.

Findings

The $D_s$-optimality criterion effectively discriminates between dependence structures.

Examples demonstrate the method's strength in selecting the correct copula class.

The approach extends equivalence theory for improved model discrimination.

Abstract

Optimum experimental design theory has recently been extended for parameter estimation in copula models. However, the choice of the correct dependence structure still requires wider analyses. In this work the issue of copula selection is treated by using discrimination design techniques. The new proposed approach consists in the use of $D_s$-optimality following an extension of corresponding equivalence theory. We also present some examples and highlight the strength of such a criterion as a way to discriminate between various classes of dependences.