Privacy-sensitive Learning for Medical Data Sharing
With Prof. Anand D. Sarwate
Department of Electrical and Computer Engineering
Rutgers, The State University of New Jersey
Faculty Host: Bobak Nazer
Refreshments will be served outside Room 339 at 1:45 p.m.
The massive explosion in digitized health data and widespread adoption of electronic medical records has the potential to revolutionize the way in which we understand healthcare. The key to unlocking this potential is sharing access to the data between institutions so that researchers can learn from large populations. However, such data sharing comes with serious privacy risks, and the current regulatory regime has resulted in a proliferation of institution-specific data-use agreements (DUAs). An alternative to data sharing is to instead share privacy-preserving data derivatives. By designing distributed algorithms that aggregate local private computations, we can learn from distributed medical data without sharing the data itself. Differential privacy is a framework for quantifying the privacy guaranteed by such procedures. In this talk I will describe this framework and how to how to design differentially-private algorithms for common machine learning tasks such as classification. I will show how these algorithms can be used to build a privacy-preserving system for learning from distributed neuroimaging data.
About the Speaker: Anand D. Sarwate joined as an Assistant Professor in the Department of Electrical and Computer Engineering at Rutgers, the State University of New Jersey in 2014. He received B.S. degrees in Electrical Engineering and Mathematics from MIT in 2002, an M.S. in Electrical Engineering from UC Berkeley in 2005 and a PhD in Electrical Engineering from UC Berkeley in 2008. From 2008-2011 he was a postdoctoral researcher at the Information Theory and Applications Center at UC San Diego and from 2011-2013 he was a Research Assistant Professor at the Toyota Technological Institute at Chicago, a philanthropically endowed academic computer science institute located on the University of Chicago campus.