This National Science Foundation (NSF) Computer and Information Science and Engineering (CISE) Project Grant award, with a total funding of $174,770, supports the development of an adaptive, federated, continuous learning system that uses a novel federated, semi-supervised learning framework. This framework aims to retrain deep neural network models on distributed, unlabeled, heterogeneous data from edge devices, while leveraging explainable AI techniques to expedite local training. The project addresses challenges related to data privacy, unlabeled heterogeneous data, and device heterogeneity in deploying AI services across diverse domains. The work will be conducted by Middle Tennessee State University and is expected to run from October 2024 to September 2026. The project also includes an educational component to involve students from groups underrepresented in computing.
Generated 8/13/24, 5:42 AM