Privacy-Preserving One-Class Support Vector Machine with Vertically Partitioned Data
Qiang Lin, Huimin Pei, Kuaini Wang, Ping Zhong · International Journal of Multimedia and Ubiquitous Engineering · 2016
We establish a new model of privacy-preserving one-class support vector machine (SVM) based on vertically partitioned data. Every participant holds all the data with a part of attributes. They apply different random matrices to establish their own kernel matrix. By sharing these partial kernel matrices, we construct a global kernel matrix and establish linear and nonlinear privacy-preserving models. Experimental results on benchmark data sets verify the validity of the proposed models.