OptiSGD-DPWGAN: Integrating Metaheuristic Algorithms and Differential Privacy to Improve Privacy-Utility Trade-Off in Generative Models

Recently, the necessity to balance model performance with data confidentiality in synthetic data generation has become a significant challenge in deep learning analysis of medical databases. In this paper, the OptiSGD-DPWGAN model is proposed, that incorporates metaheuristic algorithms and different...

Full description

Saved in:
Bibliographic Details
Main Authors: Alshaymaa Ahmed Mohamed, Yasmine N. M. Saleh, Ayman A. Abdel-Hamid
Format: Article
Language:English
Published: IEEE 2024-01-01
Series:IEEE Access
Subjects:
Online Access:https://ieeexplore.ieee.org/document/10758629/
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1850173314618621952
author Alshaymaa Ahmed Mohamed
Yasmine N. M. Saleh
Ayman A. Abdel-Hamid
author_facet Alshaymaa Ahmed Mohamed
Yasmine N. M. Saleh
Ayman A. Abdel-Hamid
author_sort Alshaymaa Ahmed Mohamed
collection DOAJ
description Recently, the necessity to balance model performance with data confidentiality in synthetic data generation has become a significant challenge in deep learning analysis of medical databases. In this paper, the OptiSGD-DPWGAN model is proposed, that incorporates metaheuristic algorithms and differential privacy into the Wasserstein Generative Adversarial Network (WGAN) architecture to protect sensitive data during the training process. The integration of Simulated Annealing and Backtracking Line Search with Stochastic Gradient Descent (SGD) optimizes the exploration of the solution space of complex parameters in non-convex deep learning models, significantly avoiding local minima. In differentially private synthetic data generation, adjusting the epsilon value critically influences the trade-off between preserving privacy and maintaining the utility of the data. Typically, a lower epsilon value strengthens privacy guarantees but can inversely affect the model’s effectiveness due to increased noise in the data processing. Empirical results demonstrate that OptiSGD-DPWGAN effectively mitigates this trade-off. Compared to other schemes, OptiSGD-DPWGAN consistently achieves lower privacy costs without compromising the quality of the synthetic data generated. These results not only show the capability of OptiSGD-DPWGAN to set a new standard in privacy-preserving synthetic data generation but also highlight its potential to generate high-quality synthetic data crucial for the medical domain which requires strict confidentiality and high precision.
format Article
id doaj-art-8b2944eb75bd48fbb31fa7ff0a9c0b08
institution OA Journals
issn 2169-3536
language English
publishDate 2024-01-01
publisher IEEE
record_format Article
series IEEE Access
spelling doaj-art-8b2944eb75bd48fbb31fa7ff0a9c0b082025-08-20T02:19:53ZengIEEEIEEE Access2169-35362024-01-011217607017608610.1109/ACCESS.2024.350290910758629OptiSGD-DPWGAN: Integrating Metaheuristic Algorithms and Differential Privacy to Improve Privacy-Utility Trade-Off in Generative ModelsAlshaymaa Ahmed Mohamed0https://orcid.org/0009-0000-2667-3148Yasmine N. M. Saleh1https://orcid.org/0000-0002-2147-8017Ayman A. Abdel-Hamid2https://orcid.org/0000-0001-9953-7147Computer Science Department, College of Computing and Information Technology, Arab Academy for Science, Technology, and Maritime Transport, Alexandria, EgyptComputer Science Department, College of Computing and Information Technology, Arab Academy for Science, Technology, and Maritime Transport, Alexandria, EgyptComputer Science Department, College of Computing and Information Technology, Arab Academy for Science, Technology, and Maritime Transport, Alexandria, EgyptRecently, the necessity to balance model performance with data confidentiality in synthetic data generation has become a significant challenge in deep learning analysis of medical databases. In this paper, the OptiSGD-DPWGAN model is proposed, that incorporates metaheuristic algorithms and differential privacy into the Wasserstein Generative Adversarial Network (WGAN) architecture to protect sensitive data during the training process. The integration of Simulated Annealing and Backtracking Line Search with Stochastic Gradient Descent (SGD) optimizes the exploration of the solution space of complex parameters in non-convex deep learning models, significantly avoiding local minima. In differentially private synthetic data generation, adjusting the epsilon value critically influences the trade-off between preserving privacy and maintaining the utility of the data. Typically, a lower epsilon value strengthens privacy guarantees but can inversely affect the model’s effectiveness due to increased noise in the data processing. Empirical results demonstrate that OptiSGD-DPWGAN effectively mitigates this trade-off. Compared to other schemes, OptiSGD-DPWGAN consistently achieves lower privacy costs without compromising the quality of the synthetic data generated. These results not only show the capability of OptiSGD-DPWGAN to set a new standard in privacy-preserving synthetic data generation but also highlight its potential to generate high-quality synthetic data crucial for the medical domain which requires strict confidentiality and high precision.https://ieeexplore.ieee.org/document/10758629/Differential privacydeep learninggenerative adversarial networkprivacy-utility tradeoffstochastic gradient descent
spellingShingle Alshaymaa Ahmed Mohamed
Yasmine N. M. Saleh
Ayman A. Abdel-Hamid
OptiSGD-DPWGAN: Integrating Metaheuristic Algorithms and Differential Privacy to Improve Privacy-Utility Trade-Off in Generative Models
IEEE Access
Differential privacy
deep learning
generative adversarial network
privacy-utility tradeoff
stochastic gradient descent
title OptiSGD-DPWGAN: Integrating Metaheuristic Algorithms and Differential Privacy to Improve Privacy-Utility Trade-Off in Generative Models
title_full OptiSGD-DPWGAN: Integrating Metaheuristic Algorithms and Differential Privacy to Improve Privacy-Utility Trade-Off in Generative Models
title_fullStr OptiSGD-DPWGAN: Integrating Metaheuristic Algorithms and Differential Privacy to Improve Privacy-Utility Trade-Off in Generative Models
title_full_unstemmed OptiSGD-DPWGAN: Integrating Metaheuristic Algorithms and Differential Privacy to Improve Privacy-Utility Trade-Off in Generative Models
title_short OptiSGD-DPWGAN: Integrating Metaheuristic Algorithms and Differential Privacy to Improve Privacy-Utility Trade-Off in Generative Models
title_sort optisgd dpwgan integrating metaheuristic algorithms and differential privacy to improve privacy utility trade off in generative models
topic Differential privacy
deep learning
generative adversarial network
privacy-utility tradeoff
stochastic gradient descent
url https://ieeexplore.ieee.org/document/10758629/
work_keys_str_mv AT alshaymaaahmedmohamed optisgddpwganintegratingmetaheuristicalgorithmsanddifferentialprivacytoimproveprivacyutilitytradeoffingenerativemodels
AT yasminenmsaleh optisgddpwganintegratingmetaheuristicalgorithmsanddifferentialprivacytoimproveprivacyutilitytradeoffingenerativemodels
AT aymanaabdelhamid optisgddpwganintegratingmetaheuristicalgorithmsanddifferentialprivacytoimproveprivacyutilitytradeoffingenerativemodels