midClustering
What is silhouette score in clustering?
Updated May 15, 2026
Short answer
Silhouette score measures how well clusters are separated and cohesive.
Deep explanation
It ranges from -1 to 1. High values indicate well-separated clusters. It considers intra-cluster distance and nearest-cluster distance.
Real-world example
Validating customer segmentation quality.
Common mistakes
- Using silhouette score alone without domain validation.
Follow-up questions
- What is a good silhouette score?