Convergence and Iteration Complexity of Policy Gradient Method for Infinite-horizon Reinforcement Learning

Mathematics

• Reinforcement Learning 100%
• Infinite Horizon 75%
• Gradient Method 74%
• Policy 65%
• Iteration 43%
• Asymptotic Convergence 19%
• Gradient 12%
• Stochastic Gradient 11%
• Ascent 10%
• Nonconvex Optimization 9%
• Global Analysis 9%
• Reward 9%
• Horizon 8%
• Rate of Convergence 6%
• Numerical Experiment 5%
• Alternatives 5%

Engineering & Materials Science

• Gradient methods 80%
• Reinforcement learning 66%
• Experiments 7%