Abstract
The problem of the nonparametric minimax estimation of an innitely smooth density at a given point, under random censorship, is considered. We establish the exact limiting behavior of the local minimax risk and propose the ecient kernel -type estimator based on the Kaplan-Meier estimator.
