Generalization in Reinforcement Learning: Safely Approximating the Value Function
A straightforward approach to the curse of dimensionality inreinforcement learning and dynamic programming is to replace the lookup table with a generalizing function approximat...