What is the elbow method used for in k-means clustering?
- A Choosing the number of clusters by looking for the point where added clusters stop reducing inertia
- B Removing outliers
- C Scaling the features
- D Measuring cluster purity
Answer
Choosing the number of clusters by looking for the point where added clusters stop reducing inertia
Silhouette score is a complementary measure. k-means also assumes roughly spherical clusters of similar size, which is often unrealistic.





