A note on the estimation of a distribution function and quantiles by a kernel method

Abstract

A distribution function is estimated by integrating a kernel estimator of the density. Quantiles are estimated by inverting the estimate of the distribution function. Second-order properties of both these estimators are studied.