Adversarial machine learning - WikipediaAdversarial machine learning is the study of the attacks on machine learning algorithms, and of the defenses against such attacks.[1] A survey from May 2020 revealed practitioners' common feeling for better protection of machine learning systems in industrial applications.[2] Machine learning techniques are mostly designed to work on specific problem sets, under the assumption that the training an
