Adaptive pointwise estimation for pure jump Lévy processes (Q500871)

This paper is concerned with adaptive kernel estimation of the Lévy density \(N(x)\) for bounded-variation pure-jump Lévy processes. In the paper the nonparametric adaptive kernel estimators of the function \(g(x)=xN(x)\) based on discrete observations are constructed and investigated under the condition that sample size tends to infinity. The proposed kernel method allows to estimate the Lévy density at a fixed point with a local adaptive choice. The method of local adaptive selection of the bandwidth is proposed. Local adaptive bandwidth selection such that the resulting adaptive estimator of \(g\) reaches the optimal minimax rate of convergence corresponding to the regularity of the function \(g\) is studied. An oracle inequality and a rate of convergence for the quadratic pointwise risk is provided. The method is easy to implement and examples and simulation results are given. The case of irregular sampling is also considered.