A note on jackknifing kernel regression function estimators (Corresp.)

Estimation of the value of a regression function at a point of continuity using a kernel-type estimator is considered and improvements by a jackknife technique are discussed. It is seen that a so-called generalized jackknife estimator asymptotically improves upon an ordinary kernel-type estimator. However, for a fixed sample size the generalized jackknife method may inflate the mean-square error.