Classification and clustering

Trains a k-nearest-neighbour classifier on a table with accuracy, recall, F1 and a confusion matrix, or runs k-means with its own initialisation and reports inertia.