English

Adversarial Label Learning

Machine Learning 2019-01-31 v3 Artificial Intelligence Machine Learning

Abstract

We consider the task of training classifiers without labels. We propose a weakly supervised method---adversarial label learning---that trains classifiers to perform well against an adversary that chooses labels for training data. The weak supervision constrains what labels the adversary can choose. The method therefore minimizes an upper bound of the classifier's error rate using projected primal-dual subgradient descent. Minimizing this bound protects against bias and dependencies in the weak supervision. Experiments on three real datasets show that our method can train without labels and outperforms other approaches for weakly supervised learning.

Keywords

Cite

@article{arxiv.1805.08877,
  title  = {Adversarial Label Learning},
  author = {Chidubem Arachie and Bert Huang},
  journal= {arXiv preprint arXiv:1805.08877},
  year   = {2019}
}

Comments

Accepted at AAAI19