Modular Neural Network and Classical Reinforcement Learning for Autonomous Robot Navigation: Inhibiting Undesirable Behaviors.
Eric A. AntoneloAlbert-Jan BaerveldtThorsteinn S. RögnvaldssonMaurício F. FigueiredoPublished in: IJCNN (2006)
Keyphrases
- modular neural network
- reinforcement learning
- autonomous robot navigation
- neural network
- multilayer perceptron
- state space
- function approximation
- reinforcement learning algorithms
- real robot
- dynamic programming
- optimal policy
- human behavior
- model free
- learning algorithm
- machine learning
- fuzzy sets
- evolutionary algorithm
- association rules
- fuzzy inference system
- action selection
- evolutionary learning
- electronic business