What is the difference between K-Means and K-Medoids?
Updated May 16, 2026
Short answer
K-Means uses centroids (means), while K-Medoids uses actual data points as cluster centers.
Deep explanation
K-Medoids is more robust to outliers because medoids are real points. K-Means is faster but sensitive to extreme values.
Real-world example
Clustering locations where outliers exist (GPS data).
Common mistakes
- Using K-Means in outlier-heavy datasets.
Follow-up questions
- Why is K-Medoids slower?
- When should you prefer it?