PrivPy: Enabling Scalable and General Privacy-Preserving Machine Learning
This addresses the need for practical and scalable privacy-preserving ML for users who must collaborate across untrusted environments, though it is incremental in building on existing secure computation techniques.
The authors tackled the problem of implementing privacy-preserving machine learning by introducing PrivPy, a framework that enables scalable collaborative computation with support for multiple secure computation engines and a Python front-end. They demonstrated its effectiveness on real-world datasets, including processing a 5000-by-1-million matrix.
We introduce PrivPy, a practical privacy-preserving collaborative computation framework, especially optimized for machine learning tasks. PrivPy provides an easy-to-use and highly compatible Python programming front-end which supports high-level array operations and different secure computation engines to allow for security assumptions and performance trade-offs. With PrivPy, programmers can write modern machine learning algorithms conveniently and efficiently in Python. We also design and implement a new efficient computation engine, with which people can use competing cloud providers to efficiently perform general arithmetics over real numbers. We demonstrate the usability and scalability of PrivPy using common machine learning models (e.g. logistic regression and convolutional neural networks) and real-world datasets (including a 5000-by-1-million matrix).