1: \begin{abstract}

2: This paper addresses the problem of regression to reconstruct functions,

3: which are observed with superimposed errors at random locations. We

4: address the problem in reproducing kernel Hilbert spaces. It is demonstrated

5: that the estimator, which is often derived by employing Gaussian random

6: fields, converges in the mean norm of the reproducing kernel Hilbert

7: space to the conditional expectation and this implies local and uniform

8: convergence of this function estimator. By preselecting the kernel,

9: the problem does not suffer from the curse of dimensionality.

10: 

11: The paper analyzes the statistical properties of the estimator. We

12: derive convergence properties and provide a conservative rate of convergence

13: for increasing sample sizes. \medskip{}

14: 

15: \noindent \textbf{Keywords:} Reproducing kernel Hilbert spaces {\tiny\textbullet} positive

16: definite functions {\tiny\textbullet}  Gramian matrix~{\tiny\textbullet}  Mercer kernels {\tiny\textbullet}  Statistical

17: learning theory

18: 

19: \noindent \textbf{Classification:} 62G05, 62G08, 62G20 

20: \end{abstract}

21: