Can machine learning be secure?

By Marco Barreno, Blaine Nelson, Russell Sears, Anthony D. Joseph, J. Doug Tygar. 2006

1,007 citations recorded by the source corpus.

Subfields: Artificial Intelligence, Computer Networks and Communications, Safety, Robustness & Interpretability, Privacy & Security, Signal Processing

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Artificial Intelligence

In Computer Networks and Communications

In Safety, Robustness & Interpretability

In Privacy & Security

In Signal Processing

Citations in the corpus

Model Inversion Attacks that Exploit Confidence Information and Basic Countermeasures (2015, background): “[3] consider such attacks on machine learning algorithms.”
Local Model Poisoning Attacks to Byzantine-Robust Federated Learning (2019, background): “Other security and privacy threats to machine learning: Adversarial examples [5, 57] aim to make a machine”
Distillation as a Defense to Adversarial Perturbations Against Deep Neural Networks (2015, background): “Attacks have been organized in taxonomies according to adversarial capabilities in [12], [41].”
Wild Patterns: Ten Years After the Rise of Adversarial Machine Learning (2017, methodology): “In 2006, in their famous paper, Barreno et al. [21] questioned the suitability of machine learning in adversarial settings from a broader perspective, cat-egorizing attacks against machine-learning algorithms both at training and at test time, and envisioning potential countermeasures to mitigate…”
Fine-Pruning: Defending Against Backdooring Attacks on Deep Neural Networks (2018, background): “Barreno et al. [7] presented a useful taxonomy for classifying different types of attacks on machine learning along three axes: whether the goal is to compromise the integrity or availability of the system, whether the attack is exploratory (gaining information about a trained model) or causative…”
Targeted Backdoor Attacks on Deep Learning Systems Using Data Poisoning (2017, background): “After presenting the details of the threat model above, we emphasize again the differences between this work and a long line of existing work studying poisoning attacks and defenses [10], [38], [12], [11], [68], [44], [9], [35], [70], [39], [61].”
ZOO: Zeroth Order Optimization Based Black-box Attacks to Deep Neural Networks without Training Substitute Models (2017, background): “The study of adversarial attacks roots in the need for understanding the robustness of state-of-the-art machine learning models [1, 2].”
Using Machine Teaching to Identify Optimal Training-Set Attacks on Machine Learners (2015, background): “Introduction The study on security threats to intelligent agents has a long history (Nelson et al. 2009; Tan, Killourhy, and Maxion 2002; Rubinstein et al. 2008; Barreno et al. 2010; Laskov and Lippmann 2010; Barreno et al. 2006; Dalvi et al. 2004; Liu and Chawla 2009; Biggio, Fumera, and Roli 2013;”
Evasion Attacks against Machine Learning at Test Time (2013, background): “To better understand the security properties of machine learning systems in adversarial settings, paradigms from security engineering and cryptography have been adapted to the machine learning field [2, 5, 14].”
Poisoning Attacks against Support Vector Machines (2012, methodology): “Poisoning attacks have been previously studied only for simple anomaly detection methods (Barreno et al., 2006; Ru-binstein et al., 2009; Kloft & Laskov, 2010).”
Manipulating Machine Learning: Poisoning Attacks and Countermeasures for Regression Learning (2018, background): “The security of machine learning has received a lot of attention in different communities (e.g., [2], [4], [14], [25], [34].”
Towards Poisoning of Deep Learning Algorithms with Back-gradient Optimization (2017, methodology): “At the same time, the variability and sophistication of cyberattacks have tremendously increased, making machine learning systems an appealing target for cybercriminals [2, 16].”
Data Poisoning Attacks on Factorization-Based Collaborative Filtering (2016, methodology): “In this paper we provide a novel technique based on stochastic gradient Langevin dynamics optimization [10] to produce malicious users that mimic normal user behaviors in order to avoid detection, while achieving attack objectives.”
Is Feature Selection Secure against Training Data Poisoning? (2015, methodology): “The worst-case scenario in which the attacker has full knowledge of the attacked system, is usually referred to as perfect knowledge case (Biggio et al., 2014; 2013b; 2012; Kloft & Laskov, 2010; Brückner et al., 2012; Huang et al., 2011; Barreno et al., 2006), and it allows one to empirically evalua”
The security of machine learning (2010, neutral): “A preliminary version of this taxonomy appears in previous work (Barreno et al. 2006).”