Byzantine-Robust Online and Offline Distributed Reinforcement Learning

Keyphrases

• Distributed Reinforcement Learning 100%
• Byzantine-robust 100%
• Offline-online 100%
• Value Iteration 66%
• Adversarial Agents 66%
• Markov Decision Process 33%
• Near-optimal 33%
• Learning Environment 33%
• Online Environment 33%
• Novel Algorithm 33%
• Multiple Agents 33%
• Optimal Sample 33%
• Central Server 33%
• Near-optimal Policy 33%
• Fake Data 33%
• Batch Size 33%
• Pessimistic Value 33%
• Optimistic Value 33%
• Robust Mean Estimation 33%
• Fake Information 33%
• Robustness Guarantees 33%
• Sample Complexity 33%
• Offline Settings 33%

Computer Science

• Reinforcement Learning 100%
• Markov Decision Process 50%
• Multiple Agents 50%