ICML 2021spotlight28 citations

Differentially Private Correlation Clustering

Mark Bun, Marek Elias, Janardhan Kulkarni

Abstract

Correlation clustering is a widely used technique in unsupervised machine learning. Motivated by applications where individual privacy is a concern, we initiate the study of differentially private correlation clustering. We propose an algorithm that achieves subquadratic additive error compared to the optimal cost. In contrast, straightforward adaptations of existing non-private algorithms all lead to a trivial quadratic error. Finally, we give a lower bound showing that any pure differentially private algorithm for correlation clustering requires additive error $\Omega$(n).

BibTeX
@InProceedings{pmlr-v139-bun21a,
  title = 	 {Differentially Private Correlation Clustering},
  author =       {Bun, Mark and Elias, Marek and Kulkarni, Janardhan},
  booktitle = 	 {Proceedings of the 38th International Conference on Machine Learning},
  pages = 	 {1136--1146},
  year = 	 {2021},
  editor = 	 {Meila, Marina and Zhang, Tong},
  volume = 	 {139},
  series = 	 {Proceedings of Machine Learning Research},
  month = 	 {18--24 Jul},
  publisher =    {PMLR},
  pdf = 	 {http://proceedings.mlr.press/v139/bun21a/bun21a.pdf},
  url = 	 {https://proceedings.mlr.press/v139/bun21a.html},
  abstract = 	 {Correlation clustering is a widely used technique in unsupervised machine learning. Motivated by applications where individual privacy is a concern, we initiate the study of differentially private correlation clustering. We propose an algorithm that achieves subquadratic additive error compared to the optimal cost. In contrast, straightforward adaptations of existing non-private algorithms all lead to a trivial quadratic error. Finally, we give a lower bound showing that any pure differentially private algorithm for correlation clustering requires additive error $\Omega$(n).}
}
Differentially Private Correlation Clustering · ICML 2021