thumbnail

N.E.A.T. PONG

Python

A demonstration of the use of the NEAT genetic algorithm in the application of playing pong.

This program uses NeuroEvolution of Augmenting Topologies (NEAT), a genetic algorithm which has a neural network play pong and uses renforcement learning to train it how to get better at the game. Upon testing, the program was able to best its human counterpart 98.2% of the time from a sample size of over 100 games. This program is based on the work of Kenneth O. Stanley and [this]("https://neat-python.readthedocs.io/en/latest/neat_overview.html") library made using his research.
2026 — Built by Kyle Levy