What is an advantage of DBSCAN over k-means?
- Finds odd shapes and flags noise
- Needs no distance metric whatsoever
- Always finds spherical clusters
- Works only with labeled data
DBSCAN grows clusters from dense regions, so it handles non-convex shapes and marks sparse points as noise.
Play AI & ML ▸