Distributed learning without distress: Privacy-preserving empirical risk minimization

Keyphrases

• Privacy-preserving 100%
• Empirical Risk Minimization 100%
• Distributed Learning 100%
• Differential Privacy 75%
• Secure Computation 75%
• Data Owner 50%
• Perturbation Method 50%
• Output Perturbation 50%
• Gradient Perturbation 50%
• Learning Environment 25%
• Popular 25%
• Learning Approaches 25%
• Global Model 25%
• Private Data 25%
• Privacy Requirements 25%
• Local Model 25%
• Local Gradient 25%
• Secure multi-party Computation 25%
• Utility Gains 25%
• Iterative Learning Algorithm 25%
• Multi-party Setting 25%
• World Data 25%
• Gradient Update 25%

Computer Science

• Privacy Preserving 100%
• Distributed Learning 100%
• Risk Minimization 100%
• Differential Privacy 75%
• Learning Algorithm 25%
• Learning Approach 25%
• Privacy Requirement 25%
• secure multiparty computation 25%