Parameters Optimization for Reinforcement Learning with Nonlinear Time-Varying Strategy by Using Uniform Experiment Design.
Tien-En LinPo-Yuan YangFu-I ChouChia-Wei ChuangJyh-Horng ChouPublished in: ISPACS (2021)
Keyphrases
- reinforcement learning
- evolutionary strategy
- parameter estimation
- design parameters
- optimal design
- maximum likelihood
- nonlinear dynamics
- nonlinear optimization
- sensitivity analysis
- global optimization
- parameter settings
- design space
- parameter values
- sequential quadratic programming
- fine tuning
- neural network
- optimization problems
- evolutionary algorithm
- multi agent
- machine learning