G. Konidaris, S. Osentoski, and P. Thomas, “Value Function Approximation in Reinforcement Learning Using the Fourier Basis”, AAAI, vol. 25, no. 1, pp. 380-385, Aug. 2011.