Report from Dagstuhl Perpectives Workshop 12371 Machine Learning Methods for Computer Security Edited by

Citations in the corpus

Wild Patterns: Ten Years After the Rise of Adversarial Machine Learning (2017, background): “Re-active approaches include: ( i ) timely detection of novel attacks, ( ii ) frequent classiﬁer retraining, and ( iii ) veriﬁcation of consistency of classiﬁer decisions against training data and ground-truth labels [40, 49].”
Towards Poisoning of Deep Learning Algorithms with Back-gradient Optimization (2017, background): “This recent research field aims at understanding the security properties of current learning algorithms, as well as at developing more secure ones [7, 16, 17].”
Is Feature Selection Secure against Training Data Poisoning? (2015, background): “As advocated in a recent workshop (Joseph et al., 2013), poisoning attacks are an emerging security threat for datadriven technologies, and could become the most relevant one in the coming years, especially in the so-called bigdata scenario dominated by data-driven technologies.”
Pattern Recognition Systems under Attack: Design Issues and Research Challenges (2014, background): “Some specific examples have been suggested by R. Lippmann’s talk at the 2013 Dagstuhl Perspectives Workshop on Machine Learning Methods for Computer Security [44], including: (i) randomizing collection of training data (collect at different timings, and locations); (ii) using difficult to reverse-en”