Composite Silhouette combines micro- and macro-averaged Silhouette scores from multiple subsampled clusterings via adaptive weighting to recover the ground-truth number of clusters more accurately than either averaging method alone.
ACM SIGMOD Record31(2002)
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.LG 1years
2026 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
Composite Silhouette: A Subsampling-based Aggregation Strategy
Composite Silhouette combines micro- and macro-averaged Silhouette scores from multiple subsampled clusterings via adaptive weighting to recover the ground-truth number of clusters more accurately than either averaging method alone.