10 Facts About DBSCAN

DBSCAN is one of the most common clustering algorithms and most cited in scientific literature.

DBSCAN has a worst-case of O, and the database-oriented range-query formulation of DBSCAN allows for index acceleration.

DBSCAN requires two parameters: e and the minimum number of points required to form a dense region .

DBSCAN executes exactly one such query for each point, and if an indexing structure is used that executes a neighborhood query in, an overall average runtime complexity of is obtained .

Spectral implementation of DBSCAN is related to spectral clustering in the trivial case of determining connected graph components — the optimal clusters with no edges cut.

For performance reasons, the original DBSCAN algorithm remains preferable to its spectral implementation.

Generalized DBSCAN is a generalization by the same authors to arbitrary "neighborhood" and "dense" predicates.

Various extensions to the DBSCAN algorithm have been proposed, including methods for parallelization, parameter estimation, and support for uncertain data.

DBSCAN is used as part of subspace clustering algorithms like PreDeCon and SUBCLU.

HDBSCAN is a hierarchical version of DBSCAN which is faster than OPTICS, from which a flat partition consisting of the most prominent clusters can be extracted from the hierarchy.