Project Grant 2349879

This Project Grant award from the National Science Foundation's (NSF) Computer and Information Science and Engineering (CISE) program (CFDA 47.070) provides $200,000.00 to the New Jersey Institute of Technology (NJIT) to develop novel techniques for analyzing spatio-temporal data in artificial intelligence of things (AIoT) systems. The project aims to advance graph signal processing and graph learning methods to improve the interpretability and efficiency of AIoT data analysis. Key research thrusts include: 1) developing topology sampling and graph neural network pruning for single-layer graph models; 2) investigating multilayer graph models for spatio-temporal data processing; 3) analyzing propagation behavior and dynamic graph evolution in AIoT systems; and 4) establishing parameter-efficient transfer learning for spatio-temporal signal processing. The project will produce new algorithms and interpretable AIoT solutions to be shared publicly, while also offering educational opportunities for students. The award has an effective date of March 1, 2025 and is scheduled for completion by February 29, 2028.