ICPR
2006
IEEE
14 years 6 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...