Locally Differentially Private Reinforcement Learning for Linear Mixture Markov Decision Processes.
Chonghua LiaoJiafan HeQuanquan GuPublished in: CoRR (2021)
Keyphrases
- markov decision processes
- differentially private
- reinforcement learning
- optimal policy
- reinforcement learning algorithms
- state space
- finite state
- policy iteration
- dynamic programming
- transition matrices
- state and action spaces
- average reward
- differential privacy
- action space
- infinite horizon
- model based reinforcement learning
- partially observable
- markov decision process
- decision theoretic planning
- function approximation
- policy evaluation
- total reward
- action sets
- function approximators
- state abstraction
- least squares
- markov decision problems
- partially observable markov decision processes
- reward function
- multi agent
- supervised learning
- linear model
- learning algorithm