Assistant Professor Mariana Raykova is the recipient of a Google Faculty Research Award to support her project entitled “Privacy Preserving Linear Regression for Big Data.” The project will address the question of how to construct machine-learning models when the input database is partitioned among several parties. Because of privacy concerns and legal restrictions, the parties cannot directly share their inputs. This is a common scenario in the cases of medical, judicial, or tax databases, which are often held by multiple institutions. Raykova and her students will focus on the computation of linear regression, which is a fundamental building block in many machine-learning algorithms. They will leverage techniques from secure computation and differential privacy in order to achieve optimal privacy guarantees and solutions that scale to databases with millions of records.