Plenary Lecture
Strong Privacy-preserving Data Collection
Professor Adam Ding
Department of Mathematics
Northeastern University
USA
E-mail: adamding@gmail.com
Abstract: Data security is an increasing critical topic in the age of big data. A strong privacy-preserving data collection method prevent privacy leakage by not allowing any party in the process access to individual level data. This eliminates the necessity of a data management center to keep the raw data. Data management centers of big organizations have been subject to increasing risk of cyber attack and data breach. We provide a framework of proving strong privacy-preservation property through analysis of candidate set cardinality and information leakage through probabilistic attack. We show that a data collection procedure using matrix masking do achieve strong privacy-preservation.
Brief Biography of the Speaker: A. Adam Ding is an Associate Professor at the Mathematics Department of Northeastern University. He received Ph.D. in Statistics from Cornell University. He has previously hold visiting faculty positions in Harvard University, University of Rochester and University of Florida. He has published over 60 research papers including publications in Journal of American Statistical Association, Journal of the Royal Statistical Society Series B, Journal of Machine Learning Research, etc. Currently he is supported by two National Science Foundation grants on cybersecurity and statistical applications.