Communication-Efficient Policy Gradient Methods for Distributed Reinforcement Learning

Mathematics

• Alternatives 8%
• Communication 60%
• Controller 25%
• Convergence Rate 11%
• Distributed Systems 15%
• Gradient 30%
• Gradient Method 74%
• Iteration 8%
• Learning 24%
• Multiagent Learning 27%
• Numerical Experiment 9%
• Optimization 10%
• Policy 65%
• Reinforcement Learning 100%
• Reward 15%

Engineering & Materials Science

• Communication 33%
• Controllers 14%
• Experiments 4%
• Gradient methods 80%
• Reinforcement learning 66%