Contains all the code for 3 projects in CS440_Intro_to_AI @ Rutgers

Fire_Spread_Bot

• Implemented Pathfinding algorithms (BFS and A* algorithms) to find the shortest path on a 2D grid
• Utilized modified cost functions to change decision-making in pathfinding

Ship_Leak_Bot

• Implemented A* pathfinding algorithm on a 2D grid
• Created a fully functioning Bayesian Network
• Utilized conditional probability, conditional factoring, and marginalization to influence decision-making in pathfinding
• Implemented local-search hill-climbing algorithm

Wire_ML

• Coded a fully function Neural Network from Scratch
• Neural Network used to classify labeled images of a set of four wires as either safe or unsafe
• Built a separate neural network to tell us which wire to cut to make an unsafe image safe
• Achieved 99.4% accuracy in the binary classification task based on 500,000 unseen images.
• Acheived 99.7% accuracy in the m-ary classification task based on 500,000 unseen images.
• Preprocessed images using computer vision and extracting information only from the intersections of wires.