Dynamic DBSCAN with Euler Tour Sequences

Seiyun Shin, Ilan Shomorony, Peter Macgregor
Proceedings of The 28th International Conference on Artificial Intelligence and Statistics, PMLR 258:1495-1503, 2025.

Abstract

We propose a fast and dynamic algorithm for Density-Based Spatial Clustering of Applications with Noise (DBSCAN) that efficiently supports online updates. Traditional DBSCAN algorithms, designed for batch processing, become computationally expensive when applied to dynamic datasets, particularly in large-scale applications where data continuously evolves. To address this challenge, our algorithm leverages the Euler Tour Trees data structure, enabling dynamic clustering updates without the need to reprocess the entire dataset. This approach preserves a near-optimal accuracy in density estimation, as achieved by the state-of-the-art static DBSCAN method (Esfandiari et al., 2021). Our method achieves an improved time complexity of $O(d \log^3(n) + \log^4(n))$ for every data point insertion and deletion, where $n$ and $d$ denote the total number of updates and the data dimension, respectively. Empirical studies also demonstrate significant speedups over conventional DBSCANs in real-time clustering of dynamic datasets, while maintaining comparable or superior clustering quality.

Cite this Paper

BibTeX
@InProceedings{pmlr-v258-shin25a, title = {Dynamic DBSCAN with Euler Tour Sequences}, author = {Shin, Seiyun and Shomorony, Ilan and Macgregor, Peter}, booktitle = {Proceedings of The 28th International Conference on Artificial Intelligence and Statistics}, pages = {1495--1503}, year = {2025}, editor = {Li, Yingzhen and Mandt, Stephan and Agrawal, Shipra and Khan, Emtiyaz}, volume = {258}, series = {Proceedings of Machine Learning Research}, month = {03--05 May}, publisher = {PMLR}, pdf = {https://raw.githubusercontent.com/mlresearch/v258/main/assets/shin25a/shin25a.pdf}, url = {https://proceedings.mlr.press/v258/shin25a.html}, abstract = {We propose a fast and dynamic algorithm for Density-Based Spatial Clustering of Applications with Noise (DBSCAN) that efficiently supports online updates. Traditional DBSCAN algorithms, designed for batch processing, become computationally expensive when applied to dynamic datasets, particularly in large-scale applications where data continuously evolves. To address this challenge, our algorithm leverages the Euler Tour Trees data structure, enabling dynamic clustering updates without the need to reprocess the entire dataset. This approach preserves a near-optimal accuracy in density estimation, as achieved by the state-of-the-art static DBSCAN method (Esfandiari et al., 2021). Our method achieves an improved time complexity of $O(d \log^3(n) + \log^4(n))$ for every data point insertion and deletion, where $n$ and $d$ denote the total number of updates and the data dimension, respectively. Empirical studies also demonstrate significant speedups over conventional DBSCANs in real-time clustering of dynamic datasets, while maintaining comparable or superior clustering quality.} }
Endnote
%0 Conference Paper %T Dynamic DBSCAN with Euler Tour Sequences %A Seiyun Shin %A Ilan Shomorony %A Peter Macgregor %B Proceedings of The 28th International Conference on Artificial Intelligence and Statistics %C Proceedings of Machine Learning Research %D 2025 %E Yingzhen Li %E Stephan Mandt %E Shipra Agrawal %E Emtiyaz Khan %F pmlr-v258-shin25a %I PMLR %P 1495--1503 %U https://proceedings.mlr.press/v258/shin25a.html %V 258 %X We propose a fast and dynamic algorithm for Density-Based Spatial Clustering of Applications with Noise (DBSCAN) that efficiently supports online updates. Traditional DBSCAN algorithms, designed for batch processing, become computationally expensive when applied to dynamic datasets, particularly in large-scale applications where data continuously evolves. To address this challenge, our algorithm leverages the Euler Tour Trees data structure, enabling dynamic clustering updates without the need to reprocess the entire dataset. This approach preserves a near-optimal accuracy in density estimation, as achieved by the state-of-the-art static DBSCAN method (Esfandiari et al., 2021). Our method achieves an improved time complexity of $O(d \log^3(n) + \log^4(n))$ for every data point insertion and deletion, where $n$ and $d$ denote the total number of updates and the data dimension, respectively. Empirical studies also demonstrate significant speedups over conventional DBSCANs in real-time clustering of dynamic datasets, while maintaining comparable or superior clustering quality.
APA
Shin, S., Shomorony, I. & Macgregor, P.. (2025). Dynamic DBSCAN with Euler Tour Sequences. Proceedings of The 28th International Conference on Artificial Intelligence and Statistics, in Proceedings of Machine Learning Research 258:1495-1503 Available from https://proceedings.mlr.press/v258/shin25a.html.

Related Material