Reinforcement learning (car demo)

Architecture & Working⚙️

• The neural network: Neural network is implemented using equations rather than layered sequential structures
• The optimizer: Optimizer works by randomly searching different parameter combinations till the car drives to the finish line.

How to use💻

• Just run the program and see how the car learns. You don't need to do anything

• Firstly, the car will try different combinations of parameters. Once it can drive to the finish line using a specific parameter combination, the car will then stop training and drive through the path repeatedly and slowly

😊If you would like to donate money for this work, please leave a comment to let me know😊

note: This project can not be considered a complete demo of reinforcement learning since it does not include mechanisms such as reward, penalty, etc. This is just a random exploration of parameters.