Nonparametric estimation of the dependence function in bivariate extreme value distributions

The paper considers the problem of estimating the dependence function of a bivariate extreme survival function with standard exponential marginals. No nparametric estimators for the dependence function are proposed and their s trong uniform convergence under suitable conditions is demonstrated. Compar isons of the proposed estimators with other estimators are made in terms of bias and mean squared error. Several real data sets from various applicati ons are used to illustrate the procedures. (C) 2001 Academic Press.