The National Science Foundation awarded a $1.2 million Project Grant to Stanford University under the Computer and Information Science and Engineering federal grant program (CFDA 47.070) for work on the Foundations of Decentralized Data Science from July 1, 2022 to June 30, 2025.
The award will support development of schemes for performing common data science tasks like analytics and inference on distributed data located across networks without collecting all data in a single location. The proposed decentralized approaches aim to optimize utility, privacy, and communication efficiency by designing messages that preserve sensitive characteristics of user data while minimizing total communication costs. The work seeks to establish theoretical benchmarks for privacy-preserving distributed computation and evaluate optimality of schemes under criteria balancing accuracy, privacy protection, and network usage. Outcomes include rigorous methods and performance standards to enable efficient and private implementation of canonical data science analyses on distributed datasets while addressing bandwidth and administrative challenges of traditional centralized approaches.