--Just press the green flag and the AI starts training. --Press [SPACE]/[CLICK] to show/hide the variables and lists. --Turbo mode ([SHIFT]+[GREEN FLAG] is recommended to let the AI train much faster. UPDATE (V2.2.1): The AI trains much faster now, it usually takes 6-8hundred tries (0.5-3m depending on your device) to find the right route.
The AI learns by trial and error. It loses points over time to let it go faster and gets bonuses while it progresses and for reaching the finish. By saving the best moves, the AI eventually discovers the right (and fastest) route. If you want a more detailed explanation, make sure to look inside the project. #AI #Pathfinder #ReinforcementLearning
