Reinforcement Learning And Dynamic Programming Using Function Approximators by Robert Babuska, Lucian Busoniu & Bart de Schutter