Texas Hold 'Em Poker AI

Poker AI utilizing reinforcement learning and fuzzy logic.

This was my second year group project. Most of the game was written in Java giving us access to Scene Builder for easy GUI creation. A small part of the game was written in Haskell with a C library sitting between the Java and Haskell code. The AI started off with some heuristics representing what is considered "good play" in poker. The AI then adapted to based upon the results of the games it played. Fuzzy logic was used to determine the strength of a hand and what the probability of each move should be. The thresholds for each move were then adjusted via the reinforcement learning.