Database Reference
In-Depth Information
Internal evaluation metrics
Common internal evaluation metrics include the WCSS we covered earlier (which is ex-
actly the K-means objective function), the Davies-Bouldin index, the Dunn Index, and the
silhouette coefficient. All these measures tend to reward clusters where elements within a
cluster are relatively close together, while elements in different clusters are relatively far
away from each other.
Note
The Wikipedia page on clustering evaluation at
http://en.wikipedia.org/wiki/
Cluster_analysis#Internal_evaluation
has more details.
