Exploiting epistemic uncertainty of the deep learning models to generate adversarial samples
Ömer Faruk Tuna, Ferhat Özgür Çatak, M. Taner Eskil
Topics & Concepts
MNIST databaseAdversarial systemComputer scienceDeep neural networksDropout (neural networks)Artificial intelligenceDeep learningMachine learningFunction (biology)Uncertainty quantificationMonte Carlo methodArtificial neural networkSampling (signal processing)MathematicsStatisticsComputer visionFilter (signal processing)BiologyEvolutionary biologyAdversarial Robustness in Machine LearningAnomaly Detection Techniques and ApplicationsAdvanced Malware Detection Techniques