Machine Learning models learn how to play arcade games

This tool uses Machine Learning to play autonomously some little arcade games. Before the algorithm is able to play autonomously, user should train it by playing a few minutes to each game. This tool doesn´t use raw pixels for making its predictions. Instead, it uses internal game data: player position, ball position, obstacles positions, etc.

• Left player is controlled by ML.
• Right player is an algorithmic player that just compares paddle position with ball position

• Racing Game