Project Description
This research project aims to develop a renewable energy forecasting model using a personalized horizontal-vertical federated learning (FL) framework. The primary objective is to enhance forecasting accuracy across solar energy sources within a neighborhood of highly diverse consumers while maintaining data privacy. The personalization aspect targets site-specific adaptation, recognizing that energy generation patterns vary by geography and infrastructure. The methodology integrates both horizontal and vertical FL. Horizontal FL enables collaboration among sites with similar feature spaces (e.g., solar farms using the same sensors), while vertical FL allows learning across organizations with different features but shared forecasting targets (e.g., combining weather data from a utility company with sensor data from a wind farm). A hybrid architecture will be designed to combine both types, and local models will be adapted using personalized layers or fine-tuning techniques. Public datasets will be partitioned to simulate decentralized data ownership. Model performance will be evaluated based on forecasting accuracy, privacy preservation, and adaptability to local patterns. The study will also explore trade-offs between personalization and global model consistency within the FL framework.
Tasks and Responsibilites
The undergraduate student participating in this project will collaborate with my PhD students and engage in both technical development and research activities. The student's tasks and responsibilities include a literature review by conducting thorough research to understand existing methodologies and findings related to federated learning. This involves reading online tutorials, academic papers, articles, and other relevant materials to grasp the current state of the field. In addition, he is responsible for assisting in setting up experiments to measure power consumption during AI training sessions. This could involve working with specialized equipment for power measurement and ensuring data accuracy. Furthermore, the student will analyze the collected data using statistical tools and software, whose role might include organizing data, performing calculations, and generating visualizations to interpret the results effectively.
Desired Qualifications
None Listed.