Safe Machine Learning and Defeating Adversarial Attacks

被引：23

作者：

Rouhani, Bita Darvish ^{[1
]}

Samragh, Mohammad ^{[2
]}

Javidi, Tara ^{[2
]}

Koushanfar, Farinaz ^{[2
]}

机构：

[1] Univ Calif San Diego, La Jolla, CA 92093 USA

[2] Univ Calif San Diego, Dept Elect & Comp Engn, La Jolla, CA 92093 USA

来源：

IEEE SECURITY & PRIVACY | 2019年 / 17卷 / 02期

基金：

美国国家科学基金会;

关键词：

16;

D O I：

10.1109/MSEC.2018.2888779

中图分类号：

TP [自动化技术、计算机技术];

学科分类号：

0812 ;

摘要：

Adversarial attacks have exposed the unreliability of machine-learning (ML) models for decision making in autonomous agents. This article discusses recent research for ML model assurance in the face of adversarial attacks.

引用

页码：31 / 38

页数：8

共 16 条

[1] [Anonymous], 2015, ROBUST CONVOLUTIONAL
[2] [Anonymous], 2017, DEEP LEARNING MODELS
[3] [Anonymous], P IEEE S SEC PRIV
[4] [Anonymous], 2017, MagNet and "efficient defenses against adversarial attacks
[5] [Anonymous], 2017, P 2017 ACM SIGSAC C
[6] [Anonymous], 2014, DEEP NEURAL NETWORK
[7] [Anonymous], 2016, DEFENSIVE DISTILLATI
[8] Huang R., 2015, Learning with a strong adversary
[9] Miyato T., 2015, Distributional smoothing with virtual adversarial training
[10] Papernot N., 2016, Practical Black-Box Attacks against Machine Learning

← 1 2 →