ERL-TD: Evolutionary Reinforcement Learning Enhanced with Truncated Variance and Distillation Mutation.
Qiuzhen LinYangfan ChenLijia MaWei-Neng ChenJianqiang LiPublished in: AAAI (2024)
Keyphrases
- reinforcement learning
- temporal difference
- genetic algorithm
- reinforcement learning algorithms
- evolutionary process
- function approximation
- temporal difference learning
- td learning
- eligibility traces
- evolutionary algorithm
- temporal difference methods
- td methods
- state space
- mutation rate
- markov decision processes
- model free
- learning algorithm
- evolutionary methods
- evaluation function
- fitness function
- policy evaluation
- evolutionary computation
- natural selection
- policy search
- reinforcement learning methods
- evolutionary optimization
- action selection
- differential evolution
- population dynamics
- population size
- function approximators
- genetic operators
- multi agent
- mutation operator
- optimal control
- reinforcement learning problems
- multi objective
- machine learning
- correlation coefficient
- dynamic programming
- multi agent reinforcement learning
- genetic algorithm ga
- neural network
- crossover operator
- standard deviation