Implement Nystrom method for better scaling behaviour

[Cdebus]

Current Spectral Clsutering algorihm is based on full distance matrix calculation and Lanczos algorithm for Eigenvalue decomposition. Whilst this provides accurate result at acceptable speed, the Nystrom method via approximating the distance matrix might yield better scaling behavious

[ClaudiaComito]

@Cdebus @Markus-Goetz is this what you're referring to?

https://scikit-learn.org/stable/modules/generated/sklearn.kernel_approximation.Nystroem.html

[github-actions[bot]]

Branch 640-Implement_Nystrom_method_for_better_scaling_behaviour created!

[mrfh92]

Implementing this idea and trying out how much it improves performance would be an interesting idea for a small/mid-size student work; although not urgent because a routine for spectral clustering already exists.

Reviewed within #1109