Cluster validation: Silhouette score, Davies-Bouldin score, Calinski-Harabasz score - [Python]
Silhouette score, Davies-Bouldin score and Calinski-Harabasz score are all internal cluster validation techniques. They can determine the goodness of clustering algorithms without external references. In the case of the Silhouette score and Calinski-Harabasz score, a higher value denotes better clustering. On the other hand, a lower value of the Davies-Bouldin score indicates better clustering. We can easily calculate these metrics using built-in packages that are available in the scikit-learn library.
GitHub address: https://github.com/randomaccess2023/M...
Helpful link: https://scikit-learn.org/stable/modul...
Silhouette score (theory): https://en.wikipedia.org/wiki/Silhoue...)
Davies-Bouldin score (theory): https://en.wikipedia.org/wiki/Davies%...
Calinski-Harabasz score: https://www.geeksforgeeks.org/calinsk...
#data_science #jupyter_notebook #cluster_validation #python #sklearn