A practical approach to evaluating the adversarial distance for machine learning classifiers

Open in new window