fr
fr

Question from the Python Machine Learning test

How are the initial centers of the clusters of the k-mean algorithm generally selected?