Enhancing Feature-Specific Data Protection via Bayesian Coordinate Differential Privacy

Maryam Aliakbarpour, Syomantak Chaudhuri, Thomas Courtade, Alireza Fallah, Michael Jordan
Proceedings of The 28th International Conference on Artificial Intelligence and Statistics, PMLR 258:4069-4077, 2025.

Abstract

Local Differential Privacy (LDP) offers strong privacy guarantees without requiring users to trust external parties. However, LDP applies uniform protection to all data features, including less sensitive ones, which degrades performance of downstream tasks. To overcome this limitation, we propose a Bayesian framework, Bayesian Coordinate Differential Privacy (BCDP), that enables feature-specific privacy quantification. This more nuanced approach complements LDP by adjusting privacy protection according to the sensitivity of each feature, enabling improved performance of downstream tasks without compromising privacy. We characterize the properties of BCDP and articulate its connections with standard non-Bayesian privacy frameworks. We further apply our BCDP framework to the problems of private mean estimation and ordinary least-squares regression. The BCDP-based approach obtains improved accuracy compared to a purely LDP-based approach, without compromising on privacy.

Cite this Paper

BibTeX
@InProceedings{pmlr-v258-aliakbarpour25a, title = {Enhancing Feature-Specific Data Protection via Bayesian Coordinate Differential Privacy}, author = {Aliakbarpour, Maryam and Chaudhuri, Syomantak and Courtade, Thomas and Fallah, Alireza and Jordan, Michael}, booktitle = {Proceedings of The 28th International Conference on Artificial Intelligence and Statistics}, pages = {4069--4077}, 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/aliakbarpour25a/aliakbarpour25a.pdf}, url = {https://proceedings.mlr.press/v258/aliakbarpour25a.html}, abstract = {Local Differential Privacy (LDP) offers strong privacy guarantees without requiring users to trust external parties. However, LDP applies uniform protection to all data features, including less sensitive ones, which degrades performance of downstream tasks. To overcome this limitation, we propose a Bayesian framework, Bayesian Coordinate Differential Privacy (BCDP), that enables feature-specific privacy quantification. This more nuanced approach complements LDP by adjusting privacy protection according to the sensitivity of each feature, enabling improved performance of downstream tasks without compromising privacy. We characterize the properties of BCDP and articulate its connections with standard non-Bayesian privacy frameworks. We further apply our BCDP framework to the problems of private mean estimation and ordinary least-squares regression. The BCDP-based approach obtains improved accuracy compared to a purely LDP-based approach, without compromising on privacy.} }
Endnote
%0 Conference Paper %T Enhancing Feature-Specific Data Protection via Bayesian Coordinate Differential Privacy %A Maryam Aliakbarpour %A Syomantak Chaudhuri %A Thomas Courtade %A Alireza Fallah %A Michael Jordan %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-aliakbarpour25a %I PMLR %P 4069--4077 %U https://proceedings.mlr.press/v258/aliakbarpour25a.html %V 258 %X Local Differential Privacy (LDP) offers strong privacy guarantees without requiring users to trust external parties. However, LDP applies uniform protection to all data features, including less sensitive ones, which degrades performance of downstream tasks. To overcome this limitation, we propose a Bayesian framework, Bayesian Coordinate Differential Privacy (BCDP), that enables feature-specific privacy quantification. This more nuanced approach complements LDP by adjusting privacy protection according to the sensitivity of each feature, enabling improved performance of downstream tasks without compromising privacy. We characterize the properties of BCDP and articulate its connections with standard non-Bayesian privacy frameworks. We further apply our BCDP framework to the problems of private mean estimation and ordinary least-squares regression. The BCDP-based approach obtains improved accuracy compared to a purely LDP-based approach, without compromising on privacy.
APA
Aliakbarpour, M., Chaudhuri, S., Courtade, T., Fallah, A. & Jordan, M.. (2025). Enhancing Feature-Specific Data Protection via Bayesian Coordinate Differential Privacy. Proceedings of The 28th International Conference on Artificial Intelligence and Statistics, in Proceedings of Machine Learning Research 258:4069-4077 Available from https://proceedings.mlr.press/v258/aliakbarpour25a.html.

Related Material