Almost sure convergence of nonparametric regression estimates (Q1059345)

Let (X,Y), \((X_ i,Y_ i)\), \(i=1,...,n\) be i.i.d. \(R^ d\times R^ 1\)-valued random vectors with \(E(| Y|)<\infty\) and \(m(x)=E(Y| X=x)\) the regression function. Select the weight functions \(W_{ni}(x)=W_{ni}(x;X_ 1,...,X_ n),\) and use \(m_ n(x)=\sum^{n}_{i=1}W_{ni}(x)Y_ i\) as an estimator of m(x). This paper shows that \(\lim_{n\to \infty}m_ n(X)=m(X)\), a.s., under weaker conditions.