Reliable validation of Reinforcement Learning Benchmarks.
Matthias Müller-BrockhausenAske PlaatMike PreussPublished in: CoRR (2022)
Keyphrases
- reinforcement learning
- machine learning
- function approximation
- cost effective
- supervised learning
- data mining
- benchmark suite
- reinforcement learning methods
- control problems
- temporal difference
- optimal policy
- state space
- evolutionary algorithm
- genetic algorithm
- learning process
- multi agent
- social networks
- real robot
- learning agents
- policy search
- real world
- robotic control