ICPR
2006
IEEE
15 years 10 months ago
2006
IEEE
To accelerate the learning of reinforcement learning, many types of function approximation are used to represent state value. However function approximation reduces the accuracy o...