keywords data mining; normalized compression distance; clustering; dendrogram; image processing; human activity recognition; silhouette coefficient; similarity