Adversarial Sequence Tagging

Abstract

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 Hamming loss. Unfortunately, this often either leads to inconsistent predictors (e.g… (More)

Topics

4 Figures and Tables

Cite this paper

@inproceedings{Li2016AdversarialST, title={Adversarial Sequence Tagging}, author={Jia Li and Kaiser Asif and Hong Wang and Brian D. Ziebart and Tanya Y. Berger-Wolf}, booktitle={IJCAI}, year={2016} }