EDM-DRL: Toward Stable Reinforcement Learning Through Ensembled Directed Mutation.
Michael H. PrinceAndrew J. McGeheeDaniel R. TauritzPublished in: EvoApplications (2021)
Keyphrases
- reinforcement learning
- genetic algorithm
- evolutionary algorithm
- function approximation
- learning algorithm
- differential evolution
- multi agent
- reinforcement learning algorithms
- temporal difference
- genetic algorithm ga
- fitness function
- state space
- model free
- markov decision processes
- machine learning
- partially observable
- robotic control
- temporal difference learning
- multi agent reinforcement learning
- neural network
- optimal control
- optimal policy
- supervised learning
- markov decision process
- evolutionary programming
- multi objective
- learning process
- clonal selection
- data mining