The Kmeans clustering and spectral clustering are two popular clustering methods for grouping similar data points together according to their similarities. However. the performance of Kmeans clustering might be quite unstable due to the random initialization of the cluster centroids. Generally. https://www.nacrack.com/