Accelerated Hierarchical Density Clustering

Leland McInnes, John Healy
5/20/2017
stat.ML

Official code: scikit-learn-contrib/hdbscan (3,136 stars, BSD-3-Clause, last commit 2026-03-27).

Abstract

We present an accelerated algorithm for hierarchical density based clustering. Our new algorithm improves upon HDBSCAN*, which itself provided a significant qualitative improvement over the popular DBSCAN algorithm. The accelerated HDBSCAN* algorithm provides comparable performance to DBSCAN, while supporting variable density clusters, and eliminating the need for the difficult to tune distance scale parameter. This makes accelerated HDBSCAN* the default choice for density based clustering. Library available at: https://github.com/scikit-learn-contrib/hdbscan

View on arXivView PDF

Code Implementations(1)

scikit-learn-contrib/hdbscanOfficial100%
BSD-3-Clause

A high performance implementation of HDBSCAN clustering.

3,136534TeX, ShellApr 22, 20156 months agoBSD-3-Clause
cluster-analysisclusteringclustering-algorithmclustering-evaluationmachine-learning+1 more

Similar Papers

Scalable DBSCAN with Random Projections

84% match2024

ParChain: A Framework for Parallel Hierarchical Agglomerative Clustering using Nearest-Neighbor Chain

83% match1 repo82021

PECANN: Parallel Efficient Clustering with Graph-Based Approximate Nearest Neighbor Search

83% match2023

Data Aggregation for Hierarchical Clustering

82% match1 repo8312023

The K-SCAN Clustering Algorithm

81% match2026

aweSOM: a CPU/GPU-accelerated Self-organizing Map and Statistically Combined Ensemble Framework for Machine-learning Clustering Analysis

81% match2025

Cite this paper

@article{mcinnes2017accelerated,
  title  = {Accelerated Hierarchical Density Clustering},
  author = {Leland McInnes and John Healy},
  year   = {2017},
  eprint = {1705.07321},
  archivePrefix = {arXiv},
  url    = {https://arxiv.org/abs/1705.07321}
}

Add a badge to your README

Shows whether official code is linked for this paper.

Code of Paper badge
[![Code of Paper](https://codeofpaper.com/badge/1705.07321.svg)](https://codeofpaper.com/paper/1705.07321)

Discussion