Dimension Independent Data Sets Approximation and Applications to Classification

Patrick Guidotti

We revisit the classical kernel method of approximation/interpolation theory in a very specific context motivated by the desire to obtain a robust procedure to approximate discrete…