rl_reach: Reproducible Reinforcement Learning Experiments for Robotic Reaching Tasks.
Pierre AumjaudDavid McAuliffeFrancisco Javier Rodríguez LeraPhilip CardiffPublished in: CoRR (2021)
Keyphrases
- reinforcement learning
- real robot
- transfer learning
- function approximation
- robotic systems
- markov decision processes
- manipulation tasks
- reinforcement learning algorithms
- model free
- optimal policy
- multi agent
- real time
- learning classifier systems
- state space
- control problems
- learning problems
- rl algorithms
- action space
- learning process
- approximate dynamic programming
- complex domains
- optimal control
- continuous state
- robot control
- multi agent reinforcement learning
- dynamic programming
- temporal difference
- mobile robot
- partially observable domains
- reinforcement learning agents
- solve complex tasks
- previously learned
- service robots
- reinforcement learning methods
- temporal difference learning
- partially observable
- learning capabilities
- action selection
- autonomous robots
- learning algorithm