Create and Run Benchmark Study of Approximators

[mmschlk]

Run an extensive benchmark of all approximators and measure the approximation quality in terms of MSE (and other loss functions).