Efficient Secure Ridge Regression from Randomized Gaussian Elimination.
Frank Blom, Niek J. Bouman, Berry Schoenmakers, Niels de Vreede · IACR Cryptology ePrint Archive · 2019
In this paper we present practical protocols for secure ridge regression. We develop the necessary secure linear algebra tools, using only basic arithmetic over prime fields. In particular, we will show how to solve linear systems of equations and compute matrix inverses efficiently, using appropriate secure random self-reductions of these problems. The distinguishing feature of our approach is that the use of secure fixed-point arithmetic is avoided entirely, while circumventing the need for secure rational reconstruction at any stage as well. In fact, in recent follow-up works, our results have already been applied and extended to several other settings.