Improving Sample Efficiency of Model-Free Algorithms for Zero-Sum Markov Games

Keyphrases

• Value Function 100%
• Sampling Efficiency 100%
• Modeling Algorithm 100%
• Model-free Algorithms 100%
• Sample Complexity 66%
• Markov Games 66%
• Popular 33%
• Best Model 33%
• Nash Equilibrium 33%
• Variance Reduction 33%
• Novel Design 33%
• Number of States 33%
• Stage-based 33%
• Markov Decision Process 33%
• Oracle 33%
• Reinforcement Learning 33%
• Monotonicity Property 33%
• Value Differences 33%
• Multi-agent Reinforcement Learning 33%
• Coarse Correlated Equilibrium 33%

Computer Science

• Function Value 100%
• Nash Equilibrium 50%
• Theoretical Study 50%
• Reinforcement Learning 50%
• Variance Reduction 50%
• Markov Decision Process 50%
• Reference Value 50%
• Multi-Agent Reinforcement Learning 50%

Mathematics

• Sample Efficiency 100%
• Reference Value 50%
• Variance Reduction Technique 50%