Mike Heddes

Jet Fighter Ai

Reinforcement LearningSep 1, 2021

The classic Atari Jet Fighter game has relatively simple game mechanics, yet determining the optimal controls against a human player is very complex. With this project, I wanted to tackle this complex control problem by creating an Artificial Intelligent agent. This agent has been trained using advanced Deep Reinforcement Learning techniques that were also used to defeat the human world champion in the game of GO, Dota 2 and StarCraft 2.

Visitors of this page are able to challenge the Agent in a game of Jet Fighter using the interactive environment above.