Title: A classification point-of-view about conditional Kendall tau
Authors: Jean-David Fermanian - Ensae-Crest (France) [presenting]
Alexis Derumigny - University of Twente (Netherlands)
Abstract: The purpose is to show how the problem of estimating conditional Kendall's tau can be rewritten as a classification task. The conditional Kendall's tau is a conditional dependence parameter which can be interpreted as a characteristic of a given pair of observations. The goal is to predict whether the pair is concordant or discordant conditionally on some covariates. We prove consistency and asymptotic normality of a family of penalized approximate maximum likelihood estimators, including the equivalent of the logit and probit regressions in our framework. Then, we detail specific algorithms adapting usual machine learning techniques, including nearest neighbors, decision trees, random forests and neural networks, to the setting of the estimation of conditional Kendall's tau. A small simulation study compares their finite sample properties. Finally, we apply all these estimators to a dataset of European stock indexes.