Kernel adjusted density estimation
Résumé
We propose and study a kernel estimator of a density in which the kernel is adapted to the data but not fixed. The smoothing procedure is followed by a location-scale transformation to reduce bias and variance. The new method naturally leads to an adaptive choice of the smoothing parameters which avoids asymptotic expansions.
Origine : Fichiers produits par l'(les) auteur(s)
Loading...