Counterfactual Off-Policy Evaluation with Gumbel-Max Structural Causal Models.
Michael OberstDavid A. SontagPublished in: CoRR (2019)
Keyphrases
- causal models
- policy evaluation
- causal reasoning
- structural model
- least squares
- reinforcement learning
- monte carlo
- temporal difference
- causal discovery
- model free
- conditional independence
- causal relationships
- directed acyclic graph
- function approximation
- variance reduction
- policy iteration
- structural equation models
- markov decision processes
- semi parametric
- partially observable markov decision processes
- machine learning
- statistical inference
- state space