> ML_LITERATURE // CAMPELLO-2013-DENSITY-BASED-CLUSTERING-BASED-ON-HIERARCHICAL-DENSITY-ESTIMATES_v1.0
Density-Based Clustering Based on Hierarchical Density Estimates
Ricardo J. G. B. Campello, Davoud Moulavi, Jörg Sander · Pacific-Asia Conference on Knowledge Discovery and Data Mining (PAKDD) (2013)
algorithm2013industry-standardthirdPartyReproduced
Principal Contribution
Extended DBSCAN to a hierarchical algorithm extracting stable clusters across varying densities without requiring an epsilon parameter.
Operational Relevance
The industry-standard choice for modern unsupervised clustering on unstructured embeddings (BERTopic, topic modeling).
Assumptions
- Clusters persist across multiple density thresholds in a minimum spanning tree based on mutual reachability distance
Limitations
- O(n^2) scaling without spatial trees; distance metric degradation in high dimensions
Connected Algorithms, Architectures & Tools
Related Algorithms:
Related Architectures:
Implementing Libraries: