Control Systems and Reinforcement Learning, Cambridge University Press, is now available!
Chapter 5 contains more on the “Q-function manifesto” discussed in this April twitter storm
Sean (@spmeyn) and I wrote a "Q-function manifesto" for the Simons Institute to share our passion for Q-functions and the linear program formulation of optimal control in Markov decision processes. Check it out and join our adventure to derive better RL algorithms! https://t.co/83jU6IX5aV
— Gergely Neu (@neu_rips) April 2, 2022