Increasing sample efficiency in deep reinforcement learning using generative environment modelling.
Per-Arne AndersenMorten GoodwinOle-Christoffer GranmoPublished in: Expert Syst. J. Knowl. Eng. (2021)
Keyphrases
- reinforcement learning
- real time
- function approximation
- generative model
- computational complexity
- data driven
- dynamic programming
- mobile robot
- unsupervised learning
- dynamic environments
- autonomous agents
- complex environments
- computing environments
- high efficiency
- real robot
- multi agent environments
- discriminative learning
- autonomous robots
- indoor environments
- computational efficiency
- state space
- multi agent
- decision trees