Variance Reduction for Reinforcement Learning in Input-Driven Environments.
Hongzi MaoShaileshh Bojja VenkatakrishnanMalte SchwarzkopfMohammad AlizadehPublished in: CoRR (2018)
Keyphrases
- variance reduction
- reinforcement learning
- gradient estimation
- sample size
- policy gradient
- monte carlo
- random numbers
- function approximation
- bias variance decomposition
- supervised learning
- confidence intervals
- importance sampling
- learning algorithm
- model selection
- naive bayes classifier
- text classification
- dynamic programming
- machine learning