Please enable JavaScript.
Coggle requires JavaScript to display documents.
Clustering Evaluation Indices (External indicators (Rand indicator (Number…
Clustering Evaluation Indices
Internal indicators
Davies-Bouldin indicator
Combines compactness and separability into one metric
Dunn indicator
Useful for data with even density and distribution
Silhouette Coefficient
Evaluates clustering between data points and average distance among different clusters
External indicators
Rand indicator
Number of true positives, true negatives, false positives, false negatives
F indicator
Considers precision and recall rate
Jaccard indicator
Measures similarity of data sets
Fowlkes - Mallows indicator
Number of true positives, true negatives, false positives and false negatives defined before
Confusion matrix
Find difference between cluster and gold standard cluster
Mutual information
Non linear correlation between clusters detected