Combination of reinforcement learning with evolution for automatically obtaining robot neural controllers.
Rodrigo Edgar Palacios-LeyvaVíctor Ricardo Cruz-ÁlvarezFernando Montes-GonzálezL. Rascon-PerezJosé Santos ReyesPublished in: IEEE Congress on Evolutionary Computation (2013)
Keyphrases
- evolutionary robotics
- real robot
- reinforcement learning
- mobile robot
- robot control
- function approximation
- neural network
- state space
- markov decision processes
- real time
- optimal control
- learning process
- robotic systems
- humanoid robot
- human robot interaction
- model free
- temporal difference
- learning algorithm
- perceptual aliasing