scientific article; zbMATH DE number 169344

Let \(X_ 1,\dots,X_ m\) i.i.d. be random vectors taking values in \(R^ n\) with a common probability density \(f(x)\). We choose the positive random variable \(H_ m(x,w)=H_ m(x,X_ 1(w),\dots,X_ m(w))\) as a random window size. The kernel estimate and the nearest neighbor estimate of \(f(x)\) are defined, respectively, as follows \[ f_ m(x)=(mH^ n_ m(x,w))^{-1}\sum^ m_{i=1} K((X_ i-x)/H_ m(x,w)), \] \[ \tilde f_ m(x)=(ma^ n_ m(x,w))^{-1}\sum^ m_{i=1} K((X_ i-x)/a_ m(x,w)),\quad m\geq 1,\quad x\in R^ n. \] Suppose that \(K(x)\) is a bounded variation function on \(R^ n\). We discuss the strong uniform consistency of \(f_ m(x)\) and \(\tilde f_ m(x)\).

Recommendations

Cited in

(6)

This page was built for publication:

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q4038019)