Browse Subject Areas

Click through the PLOS taxonomy to find articles in your field.

For more information about PLOS Subject Areas, click here.

< Back to Article

Characterizing Vocal Repertoires—Hard vs. Soft Classification Approaches

Fig 3

Comparison between the average silhouette width for K-means clustering for k = 2 to 20 clusters for all 4 feature sets.

The 9 feature set (green) shows generally higher silhouette width. For the 2-cluster-solution, all but the set based on factors (yellow) show globally the highest value. Excluding the 2-cluster-solution (not to be retained because of its lack of detail), no solution is markedly superior over all others, although plateau values of average silhouette width are already obtained for cluster numbers as small as k = 5 (apart from factor-based clustering).

Fig 3