Speaker
Description
Although government agencies, health providers, and organizations of all kinds collect and store data at an ever-increasing rate, much of the most valuable data is unavailable for research and policy due to the need to protect privacy. Extracting relevant information is especially difficult when research requires data residing in multiple locations with separate privacy controls. Stealth Software Technologies has partnered with ICPSR to implement a privacy-preserving statistical system that enables data analysis without requiring data owners to reveal any information about individuals to the analyst or to each other. Secure Multiparty Computation (MPC) is a framework that enables multiple data-holders to compute any joint function of their private inputs, while maintaining the privacy of each input. SAFRN will compute standard statistical analyses (crosstabulation, difference of means, regression, logistic regression) while provably maintaining the privacy of each participant's confidential information.
