A Dropout-Tolerated Privacy-Preserving Method for Decentralized Crowdsourced Federated Learning

Keyphrases

• Privacy-preserving 100%
• Dropout 100%
• Privacy-preserving Techniques 100%
• Crowdsourcing 100%
• Federated Learning 100%
• Identity-based 33%
• Central Server 33%
• Decrypt 33%
• Malicious Attacks 33%
• Ciphertext 33%
• Broadcast Encryption 33%
• Blockchain 16%
• Training Model 16%
• Communication Networks 16%
• Confidentiality 16%
• Still Face 16%
• Data Privacy 16%
• Computing Power 16%
• Aggregation Behavior 16%
• Computing Capabilities 16%
• Sender 16%
• Cryptographic Primitives 16%
• Homomorphic 16%
• Privacy Leakage 16%
• Data Availability 16%
• Model Accuracy 16%
• Semantically Secure 16%
• Single Point of Failure 16%
• Local Model 16%
• Model Aggregation 16%
• Training Tasks 16%
• Mobile Crowdsourcing 16%
• Decryptable 16%
• Inference Attack 16%
• Online Tasks 16%
• External Adversaries 16%
• Requester 16%

Computer Science

• Privacy Preserving 100%
• Federated Learning 100%
• Malicious Attack 28%
• Ciphertext 28%
• Blockchain 14%
• Experimental Result 14%
• Communication Network 14%
• Data Privacy 14%
• Computing Power 14%
• Cryptographic Primitive 14%
• Privacy Leakage 14%
• Data Availability 14%
• Model Accuracy 14%
• Training Model 14%
• Single Point of Failure 14%