Local linear kernel estimation of the discontinuous regression function (Q2463665)

The aim of the paper is to propose a new method for estimating a discontinuous regression function together with its jump points. The authors propose a two-step procedure that first provides estimations of the jump points and finally the method makes use of the estimated jumps to construct an estimator of the discontinuous function. The performance of the proposed method is illustrated by simulations and applied to real life data.