Differentially Private Generative Adversarial Network

Abstract

Generative Adversarial Network (GAN) and its variants have recently attracted intensive research interests due to their elegant theoretical foundation and excellent empirical performance as generative models. These tools provide a promising direction in the studies where data availability is limited. One common issue in GANs is that the density of the learned generative distribution could concentrate on the training data points, meaning that they can easily remember training samples due to the high model complexity of deep networks. This becomes a major concern when GANs are applied to private or sensitive data such as patient medical records, and the concentration of distribution may divulge critical patient information. To address this issue, in this paper we propose a differentially private GAN (DPGAN) model, in which we achieve differential privacy in GANs by adding carefully designed noise to gradients during the learning procedure. We provide rigorous proof for the privacy guarantee, as well as comprehensive empirical evidence to support our analysis, where we demonstrate that our method can generate high quality data points at a reasonable privacy level.

References in corpus (13)

• Unsupervised Representation Learning with Deep Convolutional Generative Adversarial Networks
• Deep Learning with Differential Privacy
• Stochastic Backpropagation and Approximate Inference in Deep Generative Models
• Improved Training of Wasserstein GANs
• Improved Techniques for Training GANs
• Energy-based Generative Adversarial Network
• BEGAN: Boundary Equilibrium Generative Adversarial Networks
• C-RNN-GAN: Continuous recurrent neural networks with adversarial training
• Generative Moment Matching Networks
• Generating Multi-label Discrete Patient Records using Generative Adversarial Networks
• Deep Models Under the GAN: Information Leakage from Collaborative Deep Learning
• Loss-Sensitive Generative Adversarial Networks on Lipschitz Densities
• Temporal Generative Adversarial Nets with Singular Value Clipping

Cited by in corpus (86)

• GAN-Leaks: A Taxonomy of Membership Inference Attacks against Generative Models
• Using GANs for Sharing Networked Time Series Data: Challenges, Initial Promise, and Open Questions
• How to DP-fy ML: A Practical Guide to Machine Learning with Differential Privacy
• Differentially Private Synthetic Medical Data Generation using Convolutional GANs
• FedDPGAN: Federated Differentially Private Generative Adversarial Networks Framework for the Detection of COVID-19 Pneumonia
• Privacy in Deep Learning: A Survey
• Adversarial Attacks Against Deep Generative Models on Data: A Survey
• PPGAN: Privacy-preserving Generative Adversarial Network
• Data synthesis and adversarial networks: A review and meta-analysis in cancer imaging
• Privacy-Preserving Machine Learning: Methods, Challenges and Directions
• A review of Generative Adversarial Networks for Electronic Health Records: applications, evaluation measures and data sources
• Security and Privacy Issues in Deep Learning
• Artificial Intelligence for Social Good: A Survey
• Recent Advances of Differential Privacy in Centralized Deep Learning: A Systematic Survey
• DataLens: Scalable Privacy Preserving Training via Gradient Compression and Aggregation
• No Peek: A Survey of private distributed deep learning
• Generative adversarial networks in time series: A survey and taxonomy
• Generative Models for Effective ML on Private, Decentralized Datasets
• Generating Artificial Data for Private Deep Learning
• GS-WGAN: A Gradient-Sanitized Approach for Learning Differentially Private Generators
• Synthesis of Realistic ECG using Generative Adversarial Networks
• Learning Privacy-Preserving Student Networks via Discriminative-Generative Distillation
• DECAF: Generating Fair Synthetic Data Using Causally-Aware Generative Networks
• Fed-TGAN: Federated Learning Framework for Synthesizing Tabular Data
• Differentially Private Synthetic Data: Applied Evaluations and Enhancements
• Synthetic Data: Revisiting the Privacy-Utility Trade-off
• DP-MERF: Differentially Private Mean Embeddings with Random Features for Practical Privacy-Preserving Data Generation
• Generative Adversarial Networks: A Survey Towards Private and Secure Applications
• How Faithful is your Synthetic Data? Sample-level Metrics for Evaluating and Auditing Generative Models
• BAAAN: Backdoor Attacks Against Autoencoder and GAN-Based Machine Learning Models
• Privacy attacks for automatic speech recognition acoustic models in a federated learning framework
• Measuring the quality of Synthetic data for use in competitions
• CasTGAN: Cascaded Generative Adversarial Network for Realistic Tabular Data Synthesis
• Differentially Private Generative Adversarial Networks with Model Inversion
• This Person (Probably) Exists. Identity Membership Attacks Against GAN Generated Faces
• RDP-GAN: A Rényi-Differential Privacy based Generative Adversarial Network
• ChildPredictor: A Child Face Prediction Framework with Disentangled Learning
• Relational Data Synthesis using Generative Adversarial Networks: A Design Space Exploration
• Efficient Privacy-Preserving Stochastic Nonconvex Optimization
• Representation Transfer for Differentially Private Drug Sensitivity Prediction
• Private Post-GAN Boosting
• privGAN: Protecting GANs from membership inference attacks at low cost
• Key Protected Classification for Collaborative Learning
• DTGAN: Differential Private Training for Tabular GANs
• Kamino: Constraint-Aware Differentially Private Data Synthesis
• Generative Synthesis of Insurance Datasets
• Invert and Defend: Model-based Approximate Inversion of Generative Adversarial Networks for Secure Inference
• KiNETGAN: Enabling Distributed Network Intrusion Detection through Knowledge-Infused Synthetic Data Generation
• MACE: A Flexible Framework for Membership Privacy Estimation in Generative Models
• Differentially Private Release of Israel's National Registry of Live Births
• Hide-and-Seek Privacy Challenge
• PEARL: Data Synthesis via Private Embeddings and Adversarial Reconstruction Learning
• Differentially Private Deep Learning with Direct Feedback Alignment
• Differentially Private Synthetic Data Generation Using Context-Aware GANs
• The Differentially Private Lottery Ticket Mechanism
• Differentially Private Bayesian Neural Networks on Accuracy, Privacy and Reliability
• Synthetic data shuffling accelerates the convergence of federated learning under data heterogeneity
• P3GM: Private High-Dimensional Data Release via Privacy Preserving Phased Generative Model
• FFPDG: Fast, Fair and Private Data Generation
• PATE-AAE: Incorporating Adversarial Autoencoder into Private Aggregation of Teacher Ensembles for Spoken Command Classification
• Differentially Private Variational Autoencoders with Term-wise Gradient Aggregation
• Effective and Privacy preserving Tabular Data Synthesizing
• DPD-InfoGAN: Differentially Private Distributed InfoGAN
• Foundations of Bayesian Learning from Synthetic Data
• An Efficient DP-SGD Mechanism for Large Scale NLP Models
• Obfuscation of Images via Differential Privacy: From Facial Images to General Images
• Don't Generate Me: Training Differentially Private Generative Models with Sinkhorn Divergence
• One Step to Efficient Synthetic Data
• Privacy-Enhanced Database Synthesis for Benchmark Publishing (Technical Report)
• Generative Models for Security: Attacks, Defenses, and Opportunities
• Network Generation with Differential Privacy
• Synthetic Data: Opening the data floodgates to enable faster, more directed development of machine learning methods
• A Differentially Private Multi-Output Deep Generative Networks Approach For Activity Diary Synthesis
• Private Dataset Generation Using Privacy Preserving Collaborative Learning
• Differentially Private Generative Adversarial Networks for Time Series, Continuous, and Discrete Open Data
• On the Intrinsic Differential Privacy of Bagging
• Differentially Private Survival Function Estimation
• DP-CGAN: Differentially Private Synthetic Data and Label Generation
• Integral Privacy for Sampling
• Two-Bit Aggregation for Communication Efficient and Differentially Private Federated Learning
• HydraGAN A Multi-head, Multi-objective Approach to Synthetic Data Generation
• DP-SMOTE: Integrating Differential Privacy and Oversampling Technique to Preserve Privacy in Smart Homes
• Anonymization of labeled TOF-MRA images for brain vessel segmentation using generative adversarial networks
• Learning Realistic Patterns from Unrealistic Stimuli: Generalization and Data Anonymization
• DP-SGD vs PATE: Which Has Less Disparate Impact on GANs?
• Synthetic Data Generation for Economists