Providing sequence tagging that min

Providing sequence tagging that minimize Hamming loss is a challenging, but important, task. Directly minimizing this loss over a training sample is generally an NP-hard problem. Instead, existing sequence tagging methods minimize a convex upper bound that upper bounds the Hammingloss. Unfortunately, this often either leads to inconsistent predictors (e.g., max-margin methods)or predictions that are mismatched on the Hamming loss (e.g., conditional random fields). Wepresent adversarial sequence tagging, a consistent structured prediction framework for minimizing Hamming loss by pessimistically viewing uncertainty. Our approach pessimistically approximates the training data, yielding an adversarial game between the sequence tag predictor and the sequence labeler. We demonstrate the benefits of the approach on activity recognition and information extraction/segmentation tasks.