Learn More
We describe an Adaptive Dynamic Programming algorithm VGL(λ) for learning a critic function over a large continuous state space. The algorithm, which requires a learned model of the environment,(More)