Implement the K-means clustering algorithm

[novoselrok]

Add k-means[1] clustering to doddle-model.

• Select the algorithm implementation (Lloyd's, Hartigan-Wong, ...)
• Implement initialisation methods (Random, k-means++)
• Include parallelization

[1] https://en.wikipedia.org/wiki/K-means_clustering

[inejc]

scikit-learn uses Lloyd's algorithm with k-means++ initialization and caching + the triangle inequality. Based on Wikipedia, Hartigan-Wong's method is more sophisticated and more computationally expensive. Will look more into this.

[Kamuzeriz]

K++