An Analysis of Byzantine-Tolerant Aggregation Mechanisms on Model Poisoning in Federated Learning

Keyphrases

• Aggregation Mechanism 100%
• Attacker 33%
• Backdoor 33%
• Byzantine Empire 33%
• Distributed Settings 33%
• Fairness Concern 33%
• Fixed Frequency 33%
• Frequency Attacks 33%
• Learning Settings 33%
• Malicious Attacks 33%
• Malicious Model 33%
• Malicious Participant 33%
• Model Poisoning Attack 33%
• Multiple Participants 33%

Computer Science

• Attack Scenario 33%
• Attackers 33%
• Backdoors 33%
• Exchanging Data 33%
• Federated Learning 100%
• Frequency Attack 33%
• Learning System 33%
• Machine Learning 33%
• Malicious Attack 33%
• Unique Environment 33%