What is a policy?
Action-taking policies define the logic which take in the state of the world and map it to a taken action at any moment in time.
Note that the downstream iteration of simulated ‘outcomes’ from a given action can be replaced by any kind of downstream simulation that is connected to the action iteration; we only think of it as a single ‘iterate simulation partition’ here for convenience.
References and further reading
Hardwick, Robert J. 2026a. “How do I automate taking actions with simulations?” umbralcalculations, September. https://umbralcalc.github.io/posts/how_do_i_automate_taking_actions_with_simulations.html.
Hardwick, Robert J. 2026b. “How does an action enter a simulation?” umbralcalculations, November. https://umbralcalc.github.io/posts/how_does_an_action_enter_a_simulation.html.
Sutton, Richard S., and Andrew G. Barto. 2018. Reinforcement Learning: An Introduction. 2nd ed. MIT Press.
Questions whose answers cite this one