Image Image Image




Post new topic Reply to topic  [ 2 posts ] 
Author Message
 Post subject: Bachelor Thesis, Counterfactual Regret Minimization (CFRM)
PostPosted: Thu Jan 24, 2013 9:20 pm 
Offline
Level1 member
User avatar

Posts: 47
Favourite Bot: Poki/Polaris
Hi guys,

this thesis is a bit dusty and old but I don't want it simply to collect dusk in the book shelf.
So if you are new to CFRM, this might help you a bit to get on track.
Even if you are already familiar with the concept you might be interested in
some of the Nash Equilibria statistics and playing styles.
Moreover, I like playing online poker, so I rather prefer publishing it here anonymously.
(Since I recompiled, LaTeX messed up some line breaks and split some words very stupidly,
but I am to lazy now to fix it, so deal with it^^)
Have fun with it!

Abstract:
Since the beginning of computer sciences, games have been a challenging
environment for artificial intelligence techniques. Whereas research
has made tremendous progress in playing perfect information games like
chess, poker still constitutes a difficult task due to the hidden information
of the occluded opponent cards. In this thesis one of the most successful
approaches to generate robust strategies for Heads Up Limit Hold’em will
be discussed. Counterfactual Regret Minimization computes strategies in-
dependently from an opponent model and the resulting strategies are provably
convergent to a Q-Nash Equilibrium. Since the full game of poker is
intractable, an abstraction is generated to scale down the game tree.
Consequently, the quality of the strategy highly depends on the abstraction.
Previous work mostly focused on techniques to abstract the game based on
information about the strength of a player’s card combination. However,
good poker players adjust their actions according to the texture of the board.
Recognizing board patterns is a very important concept of robust poker
strategies, some boards are more suitable for aggressive play and some give
opportunities to drawing card combinations, for instance. Therefore, two
new abstraction methods are introduced which can differentiate if a board is
drawy, i.e. if Flush and Straight Draws are possible, and if it contains low
card combinations or top ranked cards. It is shown how these new abstraction
techniques outperform the approaches that are solely based on hand
strength measures.


link


Top
 Profile E-mail  
 
 Post subject: Re: Bachelor Thesis, Counterfactual Regret Minimization (CFRM)
PostPosted: Fri Jan 25, 2013 12:50 pm 
Offline
Level1 member
User avatar

Posts: 47
Favourite Bot: Poki/Polaris
I also uploaded it
HERE


Top
 Profile E-mail  
 
Display posts from previous:  Sort by  
Post new topic Reply to topic  [ 2 posts ] 


Who is online

Users browsing this forum: No registered users and 3 guests


You cannot post new topics in this forum
You cannot reply to topics in this forum
You cannot edit your posts in this forum
You cannot delete your posts in this forum
You cannot post attachments in this forum

Search for:
Jump to: