Optimal Balance of Privacy and Utility with Differential Privacy Deep Learning Frameworks

Kotevska, Olivera; Alamudun, Folami; Stanley, Christopher

doi:10.1109/CSCI54926.2021.00141

Optimal Balance of Privacy and Utility with Differential Privacy Deep Learning Frameworks

Conference · Wed Dec 01 04:00:00 EST 2021

DOI:https://doi.org/10.1109/CSCI54926.2021.00141· OSTI ID:1875370

^[1]; ^[1]; ^[1]

ORNL

As the number of online services has increased, the amount of sensitive data being recorded is rising. Simultaneously, the decision-making process has improved by using the vast amounts of data, where machine learning has transformed entire industries. This paper addresses the development of optimal private deep neural networks and discusses the challenges associated with this task. We focus on differential privacy implementations and finding the optimal balance between accuracy and privacy, benefits and limitations of existing libraries, and challenges of applying private machine learning models in practical applications. Our analysis shows that learning rate, and privacy budget are the key factors that impact the results, and we discuss options for these settings.

Research Organization:: Oak Ridge National Laboratory (ORNL), Oak Ridge, TN (United States)

Sponsoring Organization:: USDOE

DOE Contract Number:: AC05-00OR22725

OSTI ID:: 1875370

Country of Publication:: United States

Language:: English

References (8)

Rényi Divergence and Kullback-Leibler Divergence van Erven, Tim; Harremoes, Peter IEEE Transactions on Information Theory, Vol. 60, Issue 7 https://doi.org/10.1109/TIT.2014.2320500	journal	July 2014
The Algorithmic Foundations of Differential Privacy Dwork, Cynthia; Roth, Aaron Foundations and Trends® in Theoretical Computer Science, Vol. 9, Issue 3-4 https://doi.org/10.1561/0400000042	journal	January 2013
Not one but many Tradeoffs Zhao, Benjamin Zi Hao; Kaafar, Mohamed Ali; Kourtellis, Nicolas Proceedings of the 2020 ACM SIGSAC Conference on Cloud Computing Security Workshop https://doi.org/10.1145/3411495.3421352	conference	November 2020
Differential Privacy: An Economic Method for Choosing Epsilon Hsu, Justin; Gaboardi, Marco; Haeberlen, Andreas 2014 IEEE 27th Computer Security Foundations Symposium https://doi.org/10.1109/CSF.2014.35	conference	July 2014
Deep Learning with Differential Privacy Abadi, Martin; Chu, Andy; Goodfellow, Ian CCS'16: 2016 ACM SIGSAC Conference on Computer and Communications Security, Proceedings of the 2016 ACM SIGSAC Conference on Computer and Communications Security https://doi.org/10.1145/2976749.2978318	conference	October 2016
Auditing Data Provenance in Text-Generation Models Song, Congzheng; Shmatikov, Vitaly Proceedings of the 25th ACM SIGKDD International Conference on Knowledge Discovery & Data Mining https://doi.org/10.1145/3292500.3330885	conference	July 2019
Rényi Differential Privacy Mironov, Ilya 2017 IEEE 30th Computer Security Foundations Symposium (CSF) https://doi.org/10.1109/CSF.2017.11	conference	August 2017
Stochastic gradient descent with differentially private updates Song, Shuang; Chaudhuri, Kamalika; Sarwate, Anand D. 2013 IEEE Global Conference on Signal and Information Processing https://doi.org/10.1109/GlobalSIP.2013.6736861	conference	December 2013

Similar Records

Balancing Trade-offs: Adaptive Differential Privacy in Interpretable Machine Learning Models

Conference · Thu Jul 31 20:00:00 EDT 2025 · OSTI ID:3015572

Quantum machine learning with differential privacy

Journal Article · Fri Feb 10 19:00:00 EST 2023 · Scientific Reports · OSTI ID:1924118

Optimal Balance of Privacy and Utility with Differential Privacy Deep Learning Frameworks

Citation Formats

References (8)

Similar Records

Related Subjects