Privacy-Preserving Classification of Horizontally Partitioned Data via Random Kernels

Abstract

We propose a novel privacy-preserving nonlinear support vector machine (SVM) classifier for a data matrix A whose columns represent input space features and whose individual rows are divided into groups of rows. Each group of rows belongs to an entity that is unwilling to share its rows or make them public. Our classifier is based on the concept of a reduced kernel K(A,B?) where B? is the transpose of a completely random matrix B. The proposed classifier, which is public but does not reveal the privately-held data, has accuracy comparable to that of an ordinary SVM classifier based on the entire data.