Machine Learning in Adversarial Settings

By Patrick Drew McDaniel, Nicolas Papernot, Z. Berkay Celik. 2016

143 citations recorded by the source corpus.

Subfields: Artificial Intelligence, Safety, Robustness & Interpretability

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 Safety, Robustness & Interpretability

Citations in the corpus

Wild Patterns: Ten Years After the Rise of Adversarial Machine Learning (2017, background): “This wrong start date is implicitly acknowledged in a growing number of recent papers in the area of computer security [5, 7, 10, 12–16] and computer vision [8, 9, 17], which only consider recent work on the security of deep networks against adversarial examples, almost completely ignoring previous…”
Adversarial Perturbations Against Deep Neural Networks for Malware Classification (2016, background): “Machine learning models deployed in adversarial settings therefore need to be robust to manipulations of their inputs [14].”
Transferability in Machine Learning: from Phenomena to Black-Box Attacks using Adversarial Samples (2016, background): “The existence of such a threat vector calls for the design of defensive mechanisms [17].”
Crafting adversarial input sequences for recurrent neural networks (2016, background): “The existence of this threat vector puts machine learning models at risk when deployed in potentially adversarial settings [7].”
Smart Cities: A Survey on Data Management, Security, and Enabling Technologies (2017, background): “Although solutions to some of the security vulnerabilities in adversarial machine learning have been devised [350], [351], current solutions do not completely eliminate all vulnerabilities and more research is required to find viable solutions.”
Generative adversarial networks: introduction and outlook (2017, background): “Adversarial samples [28], [29] also contain the adversarial idea.”
PassGAN: A Deep Learning Approach for Password Guessing (2017, background): “from trained models, it was recently shown that samples could be subtly modiﬁed so that they look unchanged to the human eye, but are consistently misclassiﬁed by deep learning algorithms [52], [51], [43], [8], [9], [35], [41], [30]. Several countermeasures have been proposed [53], [77]. However, th”
Fairer machine learning in the real world: Mitigating discrimination without collecting sensitive data (2017, background): “…2014; Kroll et al., 2016; Nissenbaum, 1996), in-the-wild reliability (Z ˇ liobaite _ et al., 2016); security against adversaries (Huang et al., 2011; McDaniel et al., 2016); entrenchment of inequality (boyd and Crawford, 2012; Harcourt, 2006); risks to privacy and due process (Hildebrandt and…”
Machine Learning for 5G/B5G Mobile and Wireless Communications: Potential, Limitations, and Future Directions (2019, background): “, moving the decision boundaries apart to leave fewer regions of ambiguity) [191], [192].”
Procedural Noise Adversarial Examples for Black-Box Attacks on Deep Convolutional Networks (2018, background): “It has been shown that machine learning systems are vulnerable to attacks performed at test time [2, 23, 41, 52].”
Machine Learning in IoT Security: Current Solutions and Future Challenges (2019, background): “[207] also investigated the causes of ML methods used in adversarial settings and their consequences.”
DeepFense: Online Accelerated Defense Against Adversarial Deep Learning (2017, methodology): “tes corresponding to the three attack algorithms. TABLE IV: Details of attack algorithms and mean-squared errors (MSE) for each evaluated application. The MSE is computed for normalized pixels (range [0, 1]). The average MSE over all perturbed samples is reported for each attack. Application Attack ”
Simple Black-Box Adversarial Perturbations for Deep Networks (2016, methodology): “We refer the reader to a recent survey by McDaniel et al. (2016) and references therein.”
Simple Black-Box Adversarial Attacks on Deep Neural Networks (2017, unknown): “[17] for a review of developments there.”