Poker Game builds Poker Hand Predictor!

• 

  GUI

• 

  poker game environment

• 

  making deep learning model

• 

  betting algorithm

• 

  poker game algorithm

• 

  GUI code

• 

  GUI code

• 

  running poker game

• 

  determining who won the round

• 

  test running model

• 

  library

Inspiration

Our team enjoys playing poker, so we had a passion for working on a project related to poker. The initial idea was to simply make the game environment, but we desired to challenge ourselves, so we incorporated machine learning into our project.

What it does

The program runs a poker game with two players. While two players play the game, the program collects playing data such as five cards shown and ratio of bet to the total pot and chip. These data that reflects user behavior is used to train a deep learning model that predicts players' hands from their actions during poker games. This model can be used as a part of an AI poker bot in future work.

How we built it

While two members were working on the algorithm part of poker (betting and checking hand), one member worked on building the model architecture and set environment (file) to use the algorithm and extract necessary data for training the model. The fourth member worked on researching GUI for the poker game.

Challenges we ran into

Converting code from Java to Python: two people coded in Java, two people coded in Python

Accomplishments that we're proud of

Building a poker game environment Building a program that automatically collects user data and trains a deep-learning model

What we learned

Tkinter (Python GUI)

What's next for Untitled

Based on the hand predictor model, build a complete AI poker bot.

Built With

• java
• python
• tensorflow
• tkinter