An Empirical Evaluation of the Effect of Adversarial Labels on Classifier Accuracy Estimation

Open in new window