Dustin Morrill

TL;DR: DeepStack is introduced, an algorithm for imperfect-information settings that combines recursive reasoning to handle information asymmetry, decomposition to focus computation on the relevant decision, and a form of intuition that is automatically learned from self-play using deep learning.

TL;DR: DeepStack as discussed by the authors combines recursive reasoning to handle information asymmetry, decomposition to focus computation on the relevant decision, and a form of intuition that is automatically learned from self-play using deep learning.

TL;DR: This document serves both as an overview of the code base and an introduction to the terminology, core concepts, and algorithms across the fields of reinforcement learning, computational game theory, and search.

TL;DR: This paper proposed an online learning method for minimizing regret in large extensive-form games, which learns a function approximator online to estimate the regret for choosing a particular action and uses these estimates in place of the true regrets to define a sequence of policies.

TL;DR: The exploitability descent algorithm is presented, a new algorithm to compute approximate equilibria in two-player zero-sum extensive-form games with imperfect information, by direct policy optimization against worst-case opponents, and it is proved that when following this optimization, the exploitability of a player's strategy converges asymptotically to zero.