StackedML
Practice
Labs
Questions
Models
Pricing
Sign in
Questions
/
Unsupervised Learning
/
Clustering
/
Failure cases
← Previous
Next →
162.
DBSCAN Failure in High Dimensions
medium
DBSCAN is applied to a high-dimensional dataset. What clustering failure is most likely?
A
Distance concentration makes all pairwise distances similar, undermining the density-based neighborhood definition
B
DBSCAN produces only singleton clusters in high dimensions since no points are within epsilon of each other
C
DBSCAN fails to converge in high dimensions since the epsilon parameter grows unboundedly during iteration
D
DBSCAN labels all points as core points in high dimensions since the neighborhood volume expands with dimensionality
Sign in to verify your answer
← Back to Questions