(1)
Konidaris, G.; Osentoski, S.; Thomas, P. Value Function Approximation in Reinforcement Learning Using the Fourier Basis. AAAI 2011, 25, 380-385.