Generating synthetic personal health data using conditional generative adversarial networks combining with differential privacy

Keyphrases

• Conditional Generative Adversarial Network (cGAN) 100%
• Differential Privacy 100%
• Personal Health Data 100%
• Patient Data 40%
• Synthetic Data 40%
• Imbalanced Data 40%
• Privacy-preserving 40%
• Minority Class 40%
• Data Utility 20%
• Patient Privacy 20%
• Personal Data 20%
• Personal Health 20%
• Socioeconomic 20%
• Network Model 20%
• Training Performance 20%
• Scientific Community 20%
• Synthetic Patient 20%
• Performance Measurement 20%
• Data Sampling 20%
• Data Networks 20%
• Generative Models 20%
• Additional Inputs 20%
• Privacy Concerns 20%
• Class Imbalance 20%
• Data Characteristics 20%
• Data Privacy 20%
• Training Process 20%
• Abnormal Distribution 20%
• Data Transformation 20%
• Statistical Noise 20%
• Latent Space 20%
• Legal Restrictions 20%
• Data Sparsity 20%
• Network Training 20%
• Abnormal Data 20%
• Differentially Private 20%
• Dependency between Variables 20%
• Synthetic Data Generation 20%
• Privacy Measurement 20%
• Statistical Similarity 20%
• Machine Learning Performance 20%
• Privacy Guarantees 20%
• Machine Learning Privacy 20%
• Health Datasets 20%
• Conditioning Training 20%

Computer Science

• Synthetic Data 100%
• Generative Adversarial Networks 100%
• Differential Privacy 100%
• Privacy Preserving 66%
• Minority Class 66%
• Imbalanced Data 66%
• Personal Data 33%
• Machine Learning 33%
• Data Structure 33%
• Individual Patient 33%
• Learning Performance 33%
• Privacy Concern 33%
• Data Transformation 33%
• Sparsity 33%
• Computer Network 33%
• Training Process 33%
• Generative Model 33%
• Continuous Variable 33%
• Categorical Variable 33%
• Data Characteristic 33%
• Data Generation 33%
• Request Process 33%

INIS

• data 100%
• patients 22%
• datasets 16%
• performance 11%
• diseases 5%
• solutions 5%
• space 5%
• communities 5%
• vectors 5%
• distribution 5%
• utilities 5%
• legal aspects 5%
• economics 5%
• sampling 5%
• capture 5%
• transformations 5%
• balances 5%
• noise 5%
• machine learning 5%