Computer scienceData scienceMachine learningClustering

DBSCAN

The pros and cons of DBSCAN

Report a typo

Select the statements that hold true for the advantages and disadvantages of DBSCAN:

Select one or more options from the list

Only the Euclidean metric could be used in DBSCAN

Can only produce non-spherical clusters

Capable of handling the clusters of arbitrary shapes and sizes

Does not require setting the number of clusters beforehand

DBSCAN results depend on the scale of the features

DBSCAN is sensitive to the choice of

\epsilon

and the minimum number of points in a cluster

Robust in handling the outliers

The performance degrades significantly in high-dimensional data

___

Create a free account to access the full topic