FBIT
2007
IEEE
13 years 10 months ago
2007
IEEE
The paper describes our first experiments on Reinforcement Learning to steer a real robot car. The applied method, Neural Fitted Q Iteration (NFQ) is purely data-driven based on ...