Secure PAC Bayesian Regression via Real Shamir Secret Sharing

TL;DR

This paper introduces a secure protocol for linear regression that preserves data privacy using real number secret sharing and multiparty computation, based on PAC Bayesian bounds, with two methods compared for efficiency.

Contribution

It proposes novel secure linear regression methods leveraging real number secret sharing and PAC Bayesian bounds, enhancing privacy and simplicity in distributed settings.

Findings

The secure inverse method and Gaussian elimination method are effective for privacy-preserving linear regression.

Using secret sharing on real numbers simplifies protocols and reduces communication rounds.

The leakage of information from shares is minimal and acceptable in practice.

Abstract

A common approach of system identification and machine learning is to generate a model by using training data to predict the test data instances as accurate as possible. Nonetheless, concerns about data privacy are increasingly raised, but not always addressed. We present a secure protocol for learning a linear model relying on recently described technique called real number secret sharing. We take as our starting point the PAC Bayesian bounds and deduce a closed form for the model parameters which depends on the data and the prior from the PAC Bayesian bounds. To obtain the model parameters one needs to solve a linear system. However, we consider the situation where several parties hold different data instances and they are not willing to give up the privacy of the data. Hence, we suggest to use real number secret sharing and multiparty computation to share the data and solve the linear regression in a secure way without violating the privacy of data. We suggest two methods; a secure inverse method and a secure Gaussian elimination method, and compare these methods at the end. The benefit of using secret sharing directly on real numbers is reflected in the simplicity of the protocols and the number of rounds needed. However, this comes with the drawback that a share might leak a small amount of information, but in our analysis we argue that the leakage is small.