Adversarial Machine Learning for Protecting Against Online Manipulation

Adversarial examples are inputs to a machine learning system that result in an incorrect output from system. Attacks launched through this type of input can cause severe consequences: for example, the field image recognition, stop signal be misclassified as speed limit indication. However, adversarial also represent fuel flurry research directions different domains and applications. Here, we give overview how they profitably exploited powerful tools build stronger models, capable better-withstanding attacks, two crucial tasks: fake news social bot detection.