/rl-hand-selection-poker

reinforcement learning environments for poker hand related tasks with gym and stable-baselines

Primary LanguageJupyter Notebook

Poker Hand Selection with Deep Reinforcement Learning

Train an agent to select the best five card poker hand from random cards with deep reinforcement learning. Create a custom OpenAI gym environment and then work with Stable baselines to build an agent.

  • Environments implement gym environments where the goal is to perform tasks related to five card poker hands, such as classification, selecting the best five card hand from many cards, and playing Open Face Chinese Poker

  • AgentTraining contains files for training agents in the environments

  • requirements.txt will be updated with a better way to replicate the environment used for training