Designed and implemented an autonomous AI agent capable of navigating a 2D grid environment to locate, collect, and classify waste using machine learning techniques.
- Used the A* algorithm for optimal pathfinding, with different obstacle costs.
- Developed two classification modes: a decision tree classifier and a neural network trained to categorize different types of garbage based on predefined feature sets or photos.
Technologies: Python, Pygame, scikit-learn, PyTorch
Simulation view: agent navigating the map using A* and classify detected object with cnn classifier.