Density-based spatial clustering of applications with noise (DBSCAN)

Table of Contents

Overview

  • Clustering algorithm

Preliminary

core points
A point dbscan_fefe9e556d399665a26a37824ec578cbffb0cabe.png which has at least min_pts within a distance dbscan_f774fb2b9717d65742fb7c3f1bb4e9ba2df26910.png of it (including itself) These points within distance dbscan_f774fb2b9717d65742fb7c3f1bb4e9ba2df26910.png are said to be directly reachable from dbscan_fefe9e556d399665a26a37824ec578cbffb0cabe.png.
(density-)reachable points
A point dbscan_ab437e1f9b3376761b155efe111c9860607c4b86.png is reachable from point dbscan_fefe9e556d399665a26a37824ec578cbffb0cabe.png if there exists a path dbscan_569b983f0197f100bc9ba97a2f477cd321718e9e.png with dbscan_25715787945b59d4d08de339fd19a6ce5554d127.png and dbscan_0f3ad44b3086b542c545be7eabb2e2394e33e05c.png, where each dbscan_ab768f7eab4d0651742d0ad43858f391d425d2d7.png is directly reachable from dbscan_4dfbb4efa24b820c3d05ca44f97337b43d713ff2.png. That is; all points on the path must be core points, with the possible exception of dbscan_ab437e1f9b3376761b155efe111c9860607c4b86.png.
outliers
All points not reachable from other points.