TY - GEN
T1 - SoK
T2 - 3rd IEEE European Symposium on Security and Privacy, EURO S and P 2018
AU - Papernot, Nicolas
AU - McDaniel, Patrick
AU - Sinha, Arunesh
AU - Wellman, Michael P.
N1 - Publisher Copyright:
© 2018 IEEE.
PY - 2018/7/6
Y1 - 2018/7/6
N2 - Advances in machine learning (ML) in recent years have enabled a dizzying array of applications such as data analytics, autonomous systems, and security diagnostics. ML is now pervasive - new systems and models are being deployed in every domain imaginable, leading to widespread deployment of software based inference and decision making. There is growing recognition that ML exposes new vulnerabilities in software systems, yet the technical community's understanding of the nature and extent of these vulnerabilities remains limited. We systematize findings on ML security and privacy, focusing on attacks identified on these systems and defenses crafted to date.We articulate a comprehensive threat model for ML, and categorize attacks and defenses within an adversarial framework. Key insights resulting from works both in the ML and security communities are identified and the effectiveness of approaches are related to structural elements of ML algorithms and the data used to train them. In particular, it is apparent that constructing a theoretical understanding of the sensitivity of modern ML algorithms to the data they analyze, à la PAC theory, will foster a science of security and privacy in ML.
AB - Advances in machine learning (ML) in recent years have enabled a dizzying array of applications such as data analytics, autonomous systems, and security diagnostics. ML is now pervasive - new systems and models are being deployed in every domain imaginable, leading to widespread deployment of software based inference and decision making. There is growing recognition that ML exposes new vulnerabilities in software systems, yet the technical community's understanding of the nature and extent of these vulnerabilities remains limited. We systematize findings on ML security and privacy, focusing on attacks identified on these systems and defenses crafted to date.We articulate a comprehensive threat model for ML, and categorize attacks and defenses within an adversarial framework. Key insights resulting from works both in the ML and security communities are identified and the effectiveness of approaches are related to structural elements of ML algorithms and the data used to train them. In particular, it is apparent that constructing a theoretical understanding of the sensitivity of modern ML algorithms to the data they analyze, à la PAC theory, will foster a science of security and privacy in ML.
UR - http://www.scopus.com/inward/record.url?scp=85050721582&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=85050721582&partnerID=8YFLogxK
U2 - 10.1109/EuroSP.2018.00035
DO - 10.1109/EuroSP.2018.00035
M3 - Conference contribution
AN - SCOPUS:85050721582
T3 - Proceedings - 3rd IEEE European Symposium on Security and Privacy, EURO S and P 2018
SP - 399
EP - 414
BT - Proceedings - 3rd IEEE European Symposium on Security and Privacy, EURO S and P 2018
PB - Institute of Electrical and Electronics Engineers Inc.
Y2 - 24 April 2018 through 26 April 2018
ER -