Idk what to show for seminar so i want to make a racing game where the "enemies" would be car that learned how to drive via Genetic algorithm. And i wanted to do it anyway, so 2 birds with 1 stone.

• Added Matrix Class that basically allows all matrix operations needed for neural networks, i.e Multiply Matrices, multiply be scalar, add scalar, devide, get random matrix etc...
• Added support for applying any (doube) -> (double) function on every entry of the matrix. (i.e apply activation function).
• Built a car model, track, trees, envirment in blender and added to game.
• Added Car Scripts times etc...
• Added CarAI neuralnetwork, as well as a save/load system.
• Add Sensors for AI
  • In this case simply raycast that tell the distance from "track borders" ( they are invisisble ).
• Add Genetic algorithm, and training processs.

• Possible allow AI to hit other cars in training so they can avoid collitions ( In the inital stage each CarAI has ignored all other cars ).
• Add Respawns and detect off course position.
• Possibly add Directions into the AI-Brain, try to use the Neural networks as input, and add on top of it directions.

Big thanks to the Imphenzia Youtube channel for the tutorials on Making Low Poly models, and tutorial to make a racing game. ( Reference https://www.youtube.com/watch?v=ODVV3eUE5zM&t=0s&ab_channel=Imphenzia )