1: \begin{abstract}

2: Kernel-based methods provide flexible and accurate algorithms for the reconstruction of functions from meshless samples. A major question in the use of such 

3: methods is the influence of the samples locations on the behavior of the approximation, and feasible optimal strategies are not known for general 

4: problems. 

5: 

6: Nevertheless, efficient and greedy point-selection strategies are known. This paper gives a proof of the convergence rate of the data-independent 

7: \textit{$P$-greedy} algorithm, based 

8: on the application of the convergence theory for greedy algorithms in reduced basis methods. The resulting rate of convergence is shown to be near-optimal in 

9: the case of kernels generating Sobolev spaces.

10: 

11: As a consequence, this convergence rate proves that, for kernels of Sobolev spaces, the points selected by the algorithm are asymptotically uniformly 

12: distributed, as conjectured in the paper where the algorithm has been introduced. 

13: \end{abstract}

14: