The Texas Holdem Poker game is a popular problem for both Game Theory and Artificial Intelligence. Similar to other well-known milestones like Chess and Go, poker can be partly solved by algorithms such as Tree Search, Reinforcement Learning and Nerual Networks. Nevertheless, multiplayer no-limit Texas holdem demonstrates a trickier problem, considering its vast combination of possible states and the existence of hidden information.
We built our AI mostly on Counterfactual Regret Minimization(CFR). CFR is a commonly used algorithm for poker AI. Instead of Monte Carlo CFR, which is a more popular variant of CFR, we improved our CFR algorithm with Dynamic Programming approach. That is possible because we do the sampling and the training in two different stages. By doing Sampling and Abstraction in advance, the traverse tree becomes small enough to be solved by CFR as a whole, and DP approaches become computationally feasible.
The app is implemented with our powerful and fast AI engine, and a user-friendly UI environment, with multiple configurations for you to customize your own poker game. It's handy for both training and leisure purposes.
For more details, please contact us by: [email protected]
--Terms of Use: https://www.apple.com/legal/internet-services/itunes/dev/stdeula
--Privacy Policy: http://www.sanqy-privacy-policy.sxl.cn
