The combination of deep reinforcement learning and search at both training and test time is
a powerful paradigm that has led to a number of successes in single-agent settings and …

Multiagent decision-making in partially observable environments is usually modelled as
either an extensive-form game (EFG) in game theory or a partially observable stochastic …

Imperfect-information games model strategic interactions involving multiple agents with
private information. A typical goal in this setting is to approximate an equilibrium in which all …

Many real-world situations are dynamic, with long-term interactions between multiple agents
with uncertainty and limited observations. The agents must reason about which actions to …

From the very dawn of the field, search with value functions was a fundamental concept of
computer games research. Turing's chess algorithm from 1950 was able to think two moves …

Partially observable stochastic games (POSGs) represent a very general class of models
that can be used to reason about sequential decision making in the presence of adversaries …

Limited look-ahead search with a heuristic value function is a key AI technique in
reinforcement learning and game playing. In perfectinformation problems, this has …

Many recent results in imperfect information games were only formulated for, or evaluated
on, poker-like games. We argue that sequential Bayesian games constitute a natural class of …

Search has played a fundamental role in computer game research since the very beginning.
And while online search has been commonly used in perfect information games such as …

Recent advancements in algorithms for sequential decision-making under imperfect
information have shown remarkable success in large games such as limit-and no-limit …