Neuroevolution is a Competitive Alternative to Reinforcement Learning for Skill Discovery.
Félix ChalumeauRaphael BoigeBryan LimValentin MacéMaxime AllardArthur FlajoletAntoine CullyThomas PierrotPublished in: CoRR (2022)
Keyphrases
- reinforcement learning
- function approximators
- function approximation
- artificial neural networks
- reinforcement learning algorithms
- markov decision processes
- learning algorithm
- temporal difference
- multi agent
- temporal difference learning
- real time
- model free
- robotic control
- scientific discovery
- control policy
- learning tasks
- learning capabilities
- multi agent reinforcement learning
- pattern discovery
- learning process
- lower bound
- artificial intelligence
- genetic algorithm
- machine learning
- data mining