Efficient and persistent backdoor attack by boundary trigger set constructing against federated learning

Computer Science

• Federated Learning 100%
• Backdoors 100%
• Mapping Relationship 33%
• Neural Network 16%
• Autoencoder 16%
• Training Dataset 16%
• Learning Systems 16%
• Sample Distribution 16%
• Pattern Classification 16%
• Adversarial Machine Learning 16%
• Locality Sensitive Hashing 16%
• Malicious Client 16%