From importance sampling to doubly robust policy gradient

Keyphrases

• Importance Sampling 100%
• Robust Policy 100%
• Policy Gradient 100%
• Doubly Robust Methods 100%
• Variance Reduction 60%
• Cramer-Rao Lower Bound 20%
• Flexible Forming 20%
• Finite Difference 20%
• Function Evaluation 20%
• Gradient Estimator 20%
• Simple Derivation 20%
• Doubly Robust Estimator 20%
• Off-policy Evaluation 20%
• On-policy 20%

Mathematics

• Importance Sampling 100%
• Variance Reduction 100%
• Variance 50%
• Variance Reduction Technique 50%
• Robust Estimator 50%
• Function Evaluation 50%