Skip to main content

Ai-rcade

Elise Bonner created Ai-rcade to add customized AI players to classic games. AI players are created using Tensorflow.js, Google's new machine learning library. Starting with Javascript clones of classic games, Ai-rcade uses Tensorflow.js APIs to construct neural network models — the AI players — that are trained entirely in the browser using data generated in the browser from game play.

In pong, for example, the AI model's initial training is accelerated through rapid play with a computer model. The computer player is then removed and replaced with a human player to improve the user experience. Users, in other words, do not need to train the model from scratch, and the model will continue to improve by playing against human players through inference.

Project Members: Elise Bonner

Find the program that fits your life.

Learn about our coding, cybersecurity, and data analytics bootcamps offered on full-time and part-time schedules.