Self-adapting WIP parameter setting using deep reinforcement learning.
Manuel Tomé De Andrade e SilvaAmérico AzevedoPublished in: Comput. Oper. Res. (2022)
Keyphrases
- parameter settings
- reinforcement learning
- parameter values
- parameter tuning
- function approximation
- reinforcement learning algorithms
- state space
- optimal policy
- genetic algorithm ga
- parameter optimization
- machine learning
- optimal parameters
- temporal difference
- model free
- markov decision processes
- multi agent
- learning algorithm
- real robot
- supervised learning
- parameter set
- genetic algorithm
- function approximators
- robotic control
- learning problems
- maximum likelihood
- dynamic programming
- search algorithm