Variance reduced policy evaluation with smooth function approximation

Keyphrases

• Policy Evaluation 100%
• Smooth Function 100%
• Function Approximation 100%
• Nonconvex 66%
• Neural Network 33%
• Optimization Problem 33%
• Reinforcement Learning 33%
• Stationary Point 33%
• Convergence Rate 33%
• State Action 33%
• Primal-dual 33%
• Oracle 33%
• Traditional Algorithm 33%
• Gradient Method 33%
• Approximation Function 33%
• Variance Reduction 33%
• Evaluation Challenges 33%
• Non-dual 33%
• Linear Function Approximation 33%
• Primal-dual Gradient 33%
• Two-timescale Stochastic Approximation 33%
• Nonlinear Function Approximation 33%
• Finite-sum Optimization 33%

Mathematics

• Variance 100%
• Smooth Function 100%
• Subproblem 100%
• Stationary Point 50%
• Linear Function 50%
• Variance Reduction 50%
• Approximation of Function 50%
• Finite Sum 50%
• Stochastics 50%
• Convergence Rate 50%
• Nonlinear Function 50%
• Neural Network 50%