Improving Reinforcement Learning Efficiency with Auxiliary Tasks in Non-Visual Environments: A Comparison.
Moritz LangeNoah KrystiniakRaphael C. EngelhardtWolfgang KonenLaurenz WiskottPublished in: CoRR (2023)
Keyphrases
- visual features
- reinforcement learning
- visual information
- low level
- dynamic environments
- multi agent environments
- transfer learning
- robotic systems
- real world environments
- visual tasks
- learning algorithm
- neural network
- dynamic programming
- computational complexity
- autonomous robots
- complex environments
- robot control
- computational efficiency
- previously learned
- data mining
- multi agent reinforcement learning
- working environment
- unstructured environments
- real time