Privacy-Preserving Distributed Expectation Maximization for Gaussian Mixture Model using Subspace Perturbation

09/16/2022
by   Jaron Skovsted Gundersen, et al.
0

Privacy has become a major concern in machine learning. In fact, the federated learning is motivated by the privacy concern as it does not allow to transmit the private data but only intermediate updates. However, federated learning does not always guarantee privacy-preservation as the intermediate updates may also reveal sensitive information. In this paper, we give an explicit information-theoretical analysis of a federated expectation maximization algorithm for Gaussian mixture model and prove that the intermediate updates can cause severe privacy leakage. To address the privacy issue, we propose a fully decentralized privacy-preserving solution, which is able to securely compute the updates in each maximization step. Additionally, we consider two different types of security attacks: the honest-but-curious and eavesdropping adversary models. Numerical validation shows that the proposed approach has superior performance compared to the existing approach in terms of both the accuracy and privacy level.

READ FULL TEXT

page 1

page 2

page 3

page 4

research
04/14/2021

Privacy-preserving Federated Learning based on Multi-key Homomorphic Encryption

With the advance of machine learning and the internet of things (IoT), s...
research
08/04/2021

Secure and Privacy-Preserving Federated Learning via Co-Utility

The decentralized nature of federated learning, that often leverages the...
research
05/17/2022

On the Privacy of Decentralized Machine Learning

In this work, we carry out the first, in-depth, privacy analysis of Dece...
research
11/12/2012

A Comparative Study of Gaussian Mixture Model and Radial Basis Function for Voice Recognition

A comparative study of the application of Gaussian Mixture Model (GMM) a...
research
04/29/2020

Privacy-Preserving Distributed Optimization via Subspace Perturbation: A General Framework

As the modern world becomes increasingly digitized and interconnected, d...
research
12/17/2018

Privacy-Preserving Distributed Joint Probability Modeling for Spatial-Correlated Wind Farms

Building the joint probability distribution (JPD) of multiple spatial-co...
research
09/30/2018

Privacy-preserving Stochastic Gradual Learning

It is challenging for stochastic optimizations to handle large-scale sen...

Please sign up or login with your details

Forgot password? Click here to reset