Provably Efficient Representation Learning in Low-rank Markov Decision Processes.
Weitong ZhangJiafan HeDongruo ZhouAmy ZhangQuanquan GuPublished in: CoRR (2021)
Keyphrases
- markov decision processes
- low rank
- reinforcement learning
- decision theoretic planning
- learning process
- state space
- optimal policy
- model based reinforcement learning
- partially observable
- missing data
- learning algorithm
- supervised learning
- matrix factorization
- dynamic programming
- active learning
- machine learning
- data sets
- linear combination
- higher order
- convex optimization
- reward function
- policy iteration
- rank minimization