Improved deep reinforcement learning for robotics through distribution-based experience retention.
Tim de BruinJens KoberKarl TuylsRobert BabuskaPublished in: IROS (2016)
Keyphrases
- reinforcement learning
- long term
- computer vision
- robot control
- function approximation
- state space
- artificial intelligence
- random variables
- industrial robots
- multi agent reinforcement learning
- spatial distribution
- optimal policy
- probability distribution
- learning process
- multi agent
- training data
- learning experience
- learning problems
- robotic systems
- improved algorithm
- uniformly distributed
- case study
- e learning
- perception action
- learning algorithm