Inspiration
Using probabilistic programming to structure an AI for one of the most complicated games of reading people and emotions that exists. By leveraging Bayesion networks with pymc3, we can treat variables as probability fields allowing for a lot of complex manipulation in very little code.
What it does
It analyzes the player's poker moves, as well as their face via webcam, in order to build a profile of the players moves in terms of likelihoods of doing certain actions such as bluffing. It then takes that and beats you with it!
What's next for bluffbustr
This is just the first stages of the project. The AI is complete, but can be built further to incorporate much more training data and make the models more precise and capable. The hackathon had limited testing time against only a few players so more calibration time is needed before it can reach its true potential.
Log in or sign up for Devpost to join the conversation.