Reinforcement Learning: An Introduction, Richard S. Sutton and Andrew G. Barto, 2018 (The MIT Press) - Presents a comprehensive theoretical foundation for function approximation in reinforcement learning, detailing the evolution from tabular methods to parameterized functions and introducing the application of neural networks.
Human-level control through deep reinforcement learning, Volodymyr Mnih, Koray Kavukcuoglu, David Silver, Andrei A. Rusu, Joel Veness, Marc G. Bellemare, Alex Graves, Martin Riedmiller, Andreas K. Fidjeland, Georg Ostrovski, Stig Petersen, Charles Beattie, Amir Sadik, Ioannis Antonoglou, Helen King, Dharshan Kumaran, Daan Wierstra, Shane Legg, and Demis Hassabis, 2015Nature, Vol. 518DOI: 10.1038/nature14236 - This seminal paper introduced Deep Q-Networks (DQN), demonstrating how a deep neural network could successfully approximate Q-values directly from high-dimensional sensory input, achieving human-level control in Atari games.