Improving the Performance of Batch-Constrained Reinforcement Learning in Continuous Action Domains via Generative Adversarial Networks.
Baturay SaglamOnat DalmazKaan GoncSuleyman S. KozatPublished in: SIU (2022)
Keyphrases
- reinforcement learning
- policy search
- continuous action
- continuous state
- generative model
- action space
- multi agent
- transfer learning
- reinforcement learning algorithms
- partially observable markov decision processes
- function approximation
- markov decision processes
- dynamic programming
- semi supervised
- state space
- dynamic environments
- neural network