AIfall10 /
TexasHoldem
Katlyn DeLuca
December 10, 2010
Attach:kdeluca_final_project.pdf
Overview
This project simulates Texas Hold'em hands to find a probability of each hand to win. These statistics are used and updated every time we play a new game and are used for betting.
Screenshot
Concepts Demonstrated
- bayesian Networks is used to bet based on player's hand
- Q-Learning is used to find the statistics for the chance a hand will win
Innovation
I wrote my project to keep updating the hand statistics over time (across different instances of the application)
Technology Used Block Diagram
Evaluation of Results
This project has a basic hand evaluation function set that can be improved. Hand probabilities reflect the possibility that the hand won/the times the hand was seen overall.
Additional Remarks
Setting up the framework for this project took more time than originally expected thus the hand evaluation functions are not perfect.