Chapter
Depth-Limited Search for Imperfect Information Games
13:12 - 19:23 (06:11)
The paper "Depth-Limited Search for Imperfect Information Games" introduces sound depth-limited local heads to solve imperfect information games, where learning an evaluation for a state is not enough due to dependence on both players' beliefs.
Clips
The podcast discusses the role of information abstraction and betting actions in winning at no limit Texas hold 'em.
13:12 - 15:32 (02:20)
No Limit Texas Hold 'Em
Summary
The podcast discusses the role of information abstraction and betting actions in winning at no limit Texas hold 'em. It also compares the deep learning methods used by Labradis and DeepStack in playing the game.
ChapterDepth-Limited Search for Imperfect Information Games
EpisodeTuomas Sandholm: Poker and Game Theory
PodcastLex Fridman Podcast
In this new paper, researchers have developed sound depth-limited local heads for imperfect information games, addressing the issue that the value of an information set depends on both players' beliefs and the opponent's choice of different continuation strategies.
15:33 - 19:23 (03:50)
Artificial Intelligence
Summary
In this new paper, researchers have developed sound depth-limited local heads for imperfect information games, addressing the issue that the value of an information set depends on both players' beliefs and the opponent's choice of different continuation strategies.