Avoid Overfitting in Deep Reinforcement Learning: Increasing Robustness Through Decentralized Control.
Malte SchillingPublished in: ICANN (4) (2021)
Keyphrases
- decentralized control
- avoid overfitting
- reinforcement learning
- markov decision processes
- multiagent systems
- collective behavior
- smart grid
- goal oriented
- multi agent systems
- virtual communities
- training data
- grid systems
- model free
- modular robots
- multi agent
- support vector machine
- decision trees
- power system
- decision making
- information systems
- learning algorithm
- network security
- data sets
- reinforcement learning algorithms
- vision system
- decision makers
- state space
- exchange information
- genetic algorithm