Button Menu

Computer Science Department

UNO from an AI Perspective

Randan Harris

The goal behind this project will be to study how Artificial Intelligence can be used to learn to work in a simple environment. This specific project will use the game UNO as an environment in which the AI will be able to learn. As UNO is generally a four player game, the AI will earn how to play in an environment where it must play against other AI's, humans, or a mix of both. The goal of this project is to create an Artificial Intelligence that will be able to assess the best moves for any given situation throughout the game and also be able to be competitive not only against other AI but humans as well. The AI will know how many total human and AI players as the human will be the one to set this option before the game starts.

As there are many dynamics that go into place when playing UNO, the AI will have to adapt to any given situation that may arise during gameplay. In accordance with this concept, the AI will also learn to challenge certain calls that take place within the game. This will include telling someone that they have not called UNO or deciding what color choice to pick when playing a wild card. Essentially, I will have the AI learn to be able to counter moves that it could deem advantageous to other players thus making it take control of the game itself.