On Instrumental Variable Regression for Deep Offline Policy Evaluation.
Yutian ChenLiyuan XuÇaglar GülçehreTom Le PaineArthur GrettonNando de FreitasArnaud DoucetPublished in: CoRR (2021)
Keyphrases
- policy evaluation
- semi parametric
- least squares
- temporal difference
- monte carlo
- linear regression
- reinforcement learning
- model free
- regression model
- markov decision processes
- policy iteration
- regression problems
- model selection
- function approximation
- variance reduction
- optimal policy
- evaluation function
- linear model
- support vector
- fixed point
- cost function
- statistical inference
- learning algorithm