arXiv (Cornell University) · 2019 · 35 citations · 10 references
Open access
ComputingComputer ScienceCybersecurityComputer SecurityFeature EngineeringMachine LearningDeep LearningExplainable Machine LearningThreat DetectionThreat ModelingAdversarial ExamplesExplainabilityMachine-learning AttacksSecure Machine LearningSecurity Evaluation
We present secml, an open-source Python library for secure and explainable machine learning. It implements the most popular attacks against machine learning, including not only test-time evasion attacks to generate adversarial examples against deep neural networks, but also training-time poisoning attacks against support vector machines and many other algorithms. These attacks enable evaluating the security of learning algorithms and of the corresponding defenses under both white-box and black-box threat models. To this end, secml provides built-in functions to compute security evaluation curves, showing how quickly classification performance decreases against increasing adversarial perturbations of the input data. secml also includes explainability methods to help understand why adversarial attacks succeed against a given model, by visualizing the most influential features and training prototypes contributing to each decision. It is distributed under the Apache License 2.0, and hosted at this https URL.
10
Marco Túlio Ribeiro, Sameer Singh, Carlos Guestrin · 2016 · 15.2K citations
The Limitations of Deep Learning in Adversarial Settings
Nicolas Papernot, Patrick McDaniel, Somesh Jha et al. · 2016 · 3.9K citations
Axiomatic Attribution for Deep Networks
Mukund Sundararajan, Ankur Taly, Qiqi Yan · arXiv (Cornell University) · 2017 · 2.6K citations · Full text
Understanding Black-box Predictions via Influence Functions
Pang Wei Koh, Percy Liang · arXiv (Cornell University) · 2017 · 1.2K citations · Full text
Ling Huang, Anthony D. Joseph, Blaine Nelson et al. · 2011 · 1.2K citations