I am interested in artificial intelligence and reinforcement learning, particularly in how these fields can intersect to solve real-world problems. My MASc thesis, "Robust Reinforcement Learning for Linear Temporal Logic Specifications with Finite Trajectory Duration," received the best paper award at the 37th Canadian Conference on AI! You can check it out here.
Using a combonation of Monte-Carlo Tree-Search and Neural Networks, inspired by the famous AlphaGo Zero framework, I developed a way to satisfy complex LTLf tasks.
The dashboard allows users to interact with the data, visualize solar panels on satellite images, and gain insights on potential energy output, panel efficiency, and more.
In addition to detection, we developed various machine learning models to predict solar panel energy production for a variety of sites and different future time steps.