Existing adversarial learning algorithms differ in design assumptions regarding adversary’s knowledge, attack strategies, attack influence, and security violation. In this book provides insights on the relation between adversarial learning and cybersecurity. The authors survey and summarize non-stationary data representations learnt by deep learning networks in big data, evolutionary computing, fog computing, cyber-physical systems, transfer learning, sparse learning, robust learning, and reinforcement learning. The robustness of deep learning networks is examined to produce a taxonomy of adversarial examples and algorithms. The authors also survey the use of game theory, convex optimization and stochastic optimization in adversarial deep learning formulations. weiterlesen