Federal Grant Award Summary

The National Science Foundation's Division of Mathematical Sciences awarded a Project Grant of $180,000 to the University of Wisconsin–Madison, effective July 1, 2026 through June 30, 2029, under the Mathematical and Physical Sciences (CFDA 47.049) program. This award supports fundamental research into high-dimensional asymptotics of estimation under privacy and computational constraints. The project delivers theoretical advances and algorithmic innovations addressing the critical trade-offs between statistical accuracy, privacy preservation, and computational efficiency in machine learning systems. Key deliverables include new mathematical theories for differential privacy techniques, precise characterizations of privacy loss in commonly used methods such as differentially private principal component analysis, and analysis of efficient algorithms for low-rank matrix estimation and denoising. The research aims to establish mathematically sound foundations for trustworthy, scalable, and privacy-conscious artificial intelligence systems applicable to sensitive data domains including genomics, neuroscience, healthcare, and social sciences.

No subawards are planned for this project. The research will be conducted at the University of Wisconsin–Madison campus in Madison, Wisconsin, and is expected to advance the nation's scientific enterprise by providing rigorous theoretical frameworks that enable responsible deployment of machine learning algorithms operating on high-dimensional, privacy-sensitive datasets at scale.