Actor Critic with Differentially Private Critic.
Jonathan LebensoldWilliam L. HamiltonBorja BalleDoina PrecupPublished in: CoRR (2019)
Keyphrases
- differentially private
- actor critic
- reinforcement learning
- approximate dynamic programming
- temporal difference
- optimal control
- policy gradient
- neuro fuzzy
- gradient method
- reinforcement learning algorithms
- differential privacy
- policy iteration
- function approximation
- dynamic programming
- least squares
- action selection
- model free
- linear program
- average reward
- markov decision processes
- monte carlo
- state space
- evaluation function
- dynamical systems
- learning algorithm