Free Access

Table 1.

Algorithms considered for inclusion by this study.

Runtime Deals with Open-
Algorithm scaling (a) noise source
KMeans n No sklearn(b)
Affinity propagation n2 No sklearn(b)
Mean-shift n2 No sklearn(b)
Spectral n3 No sklearn(b)
Ward n3 No sklearn(b)
Agglomerative n3 No sklearn(b)
DBSCAN nlog n Yes sklearn(b)
OPTICS n2 Yes sklearn(b)
Gaussian mixtures n No sklearn(b)
Birch n No sklearn(b)
Friend of Friends nlog n No pyfof(c)
HDBSCAN nlog n Yes HDBSCAN(d)

Notes.

(a)

Runtime scalings are best case estimates and are only given with respect to number of data points n.

Current usage metrics show cumulative count of Article Views (full-text article views including HTML views, PDF and ePub downloads, according to the available data) and Abstracts Views on Vision4Press platform.

Data correspond to usage on the plateform after 2015. The current usage metrics is available 48-96 hours after online publication and is updated daily on week days.

Initial download of the metrics may take a while.