Research Article
A Novel Reinforcement Learning Architecture for Continuous State and Action Spaces
Figure 6
Learning curves of the SARSA algorithm using different numbers of multilayer perceptrons and 10 radial basis functions for each multilayer perceptron.
