Secure and Private Machine Learning

Grant number: DE160100584 | Funding period: 2016 - 2018

Completed

Abstract

This project intends to answer the question: How can machines learn from data when participants behave maliciously for personal gain? Machine learning and statistics are used in many technologies where participants have an incentive to game the system (eg internet ad placement, e-commerce rating systems, credit risk in finance, health analytics and smart utility grids). However, little is known about how well state-of-the-art statistical inference techniques fare when data is manipulated by a malicious participant. The project's outcomes aim to ensure that statistical analysis is accurate while preserving data privacy, providing theoretical foundations of secure machine learning in adversari..

View full description

Related publications (9)

Scholarly works icon

Attacking Data Transforming Learners at Training Time

Scott Alfeld, Ara Vartanian, Lucas Newman-Johnson, Benjamin IP Rubinstein

2019-07-23

While machine learning systems are known to be vulnerable to data-manipulation attacks at both training and deployment time, littl..

Scholarly works icon

Differentially-Private Two-Party Egocentric Betweenness Centrality

L Roohi, BIP Rubinstein, V Teague

2019-04-01

© 2019 IEEE. We describe a novel protocol for computing the egocentric betweenness centrality of a node when relevant edge informa..

Scholarly works icon

Differentially private counting of users' spatial regions

Maryam Fanaeepour, Benjamin IP Rubinstein

2018-01-01

Mining of spatial data is an enabling technology for mobile services, Internet-connected cars and the Internet of Things. But the ..

Scholarly works icon

Differential Privacy for Bayesian Inference through Posterior Sampling

Christos Dimitrakakis, Blaine Nelson, Zuhe Zhang, Aikaterini Mitrokotsa, Benjamin IP Rubinstein

2017-01-01

Differential privacy formalises privacy-preserving mechanisms that provide access to a database. Can Bayesian inference be used di..

Scholarly works icon

The bernstein mechanism: Function release under differential privacy

F Aldà, BIP Rubinstein

2017-01-01

© Copyright 2017, Association for the Advancement of Artificial Intelligence (www.aaai.org). All rights reserved. We address the p..

Scholarly works icon

Beyond Points and Paths: Counting Private Bodies

M Fanaeepour, B RUBINSTEIN

2016-01-01

Mining of spatial data is an enabling technology for mobile services, Internet-connected cars, and the Internet of Things. But the..

Scholarly works icon

Large-scale strategic games and adversarial machine learning

T Alpcan, BIP Rubinstein, C Leckie

2016-01-01

Decision making in modern large-scale and complex systems such as communication networks, smart electricity grids, and cyber-physi..

Scholarly works icon

On the Differential Privacy of Bayesian Inference

Z Zhang, B RUBINSTEIN, C Dimitrakakis

2016-01-01

We study how to communicate findings of Bayesian inference to third parties, while preserving the strong guarantee of differential..

Scholarly works icon

Robust and Private Bayesian Inference

C Dimitrakakis, B Nelson, A Mitrokotsa, B RUBINSTEIN

2014-10-01

We examine the robustness and privacy of Bayesian inference, under assumptions on the prior, and with no modifications to the Baye..

Researchers