Comment by hidden-markov

The adversarial example attacks work against most known machine learning models, including statistical models, decision trees, feed-forward nets, convolutional nets, recurrent nets. They not only work against classifiers, but also against regression, and RL models.