Policy space diversity for non-transitive games
Abstract Policy-Space Response Oracles (PSRO) is an influential algorithm framework for
approximating a Nash Equilibrium (NE) in multi-agent non-transitive games. Many previous …
approximating a Nash Equilibrium (NE) in multi-agent non-transitive games. Many previous …
A survey of decision making in adversarial games
In many practical applications, such as poker, chess, drug interdiction, cybersecurity, and
national defense, players often have adversarial stances, ie, the selfish actions of each …
national defense, players often have adversarial stances, ie, the selfish actions of each …
Learning in games: a systematic review
RJ Qin, Y Yu - Science China Information Sciences, 2024 - Springer
Game theory studies the mathematical models for self-interested individuals. Nash
equilibrium is arguably the most central solution in game theory. While finding the Nash …
equilibrium is arguably the most central solution in game theory. While finding the Nash …
Perfectdou: Dominating doudizhu with perfect information distillation
As a challenging multi-player card game, DouDizhu has recently drawn much attention for
analyzing competition and collaboration in imperfect-information games. In this paper, we …
analyzing competition and collaboration in imperfect-information games. In this paper, we …
Escher: Eschewing importance sampling in games by computing a history value function to estimate regret
Recent techniques for approximating Nash equilibria in very large games leverage neural
networks to learn approximately optimal policies (strategies). One promising line of research …
networks to learn approximately optimal policies (strategies). One promising line of research …
A robust and opponent-aware league training method for StarCraft II
R Huang, X Wu, H Yu, Z Fan, H Fu… - Advances in Neural …, 2024 - proceedings.neurips.cc
It is extremely difficult to train a superhuman Artificial Intelligence (AI) for games of similar
size to StarCraft II. AlphaStar is the first AI that beat human professionals in the full game of …
size to StarCraft II. AlphaStar is the first AI that beat human professionals in the full game of …
Quality-similar diversity via population based reinforcement learning
Diversity is a growing research topic in Reinforcement Learning (RL). Previous research on
diversity has mainly focused on promoting diversity to encourage exploration and thereby …
diversity has mainly focused on promoting diversity to encourage exploration and thereby …
Game-theoretic robust reinforcement learning handles temporally-coupled perturbations
Deploying reinforcement learning (RL) systems requires robustness to uncertainty and
model misspecification, yet prior robust RL methods typically only study noise introduced …
model misspecification, yet prior robust RL methods typically only study noise introduced …
An efficient deep reinforcement learning algorithm for solving imperfect information extensive-form games
One of the most popular methods for learning Nash equilibrium (NE) in large-scale imperfect
information extensive-form games (IIEFGs) is the neural variants of counterfactual regret …
information extensive-form games (IIEFGs) is the neural variants of counterfactual regret …
Greedy when sure and conservative when uncertain about the opponents
We develop a new approach, named Greedy when Sure and Conservative when Uncertain
(GSCU), to competing online against unknown and nonstationary opponents. GSCU …
(GSCU), to competing online against unknown and nonstationary opponents. GSCU …