Estimation non-paramétrique de la fonction de répartition Bivariée

Abstract

In several areas of statistical application, the estimation of the bivariate distribution function plays an essential role, We propose such estimator based on Bernstein’s approach in the presence of right censored data.The results of the asymptotic behavior of our estimator have been established, Almost sure convergence, bias as well as asymptotic normality. We derived a nonparametric estimator of the copula from our estimator and showed the result of its convergence. Finally, we illustrated the robustness of our contribution through simulated and real data.