2017/02/28 by de Berg, Mark, Gunawan, Ade, Roeloffzen, Marcel · 1 citation
#Computational Geometry (cs.CG) #FOS: Computer and information sciences
paper · doi:10.48550/arxiv.1702.08607
We present a new algorithm for the widely used density-based clustering method DBscan. Our algorithm computes the DBscan-clustering in O(nlog n) time in ℝ2, irrespective of the scale parameter ε (and assuming the second parameter MinPts is set to a fixed constant, as is the case in practice). Experiments show that the new algorithm is not only fast in theory, but that a slightly simplified version is competitive in practice and much less sensitive to the choice of ε than the original DBscan algorithm. We also present an O(nlog n) randomized algorithm for HDBscan in the plane---HDBscan is a hierarchical version of DBscan introduced recently---and we show how to compute an approximate version of HDBscan in near-linear time in any fixed dimension.