Corpus ID: 15334003

Label Efficient Learning by Exploiting Multi-Class Output Codes

@inproceedings{Balcan2015LabelEL,
  title={Label Efficient Learning by Exploiting Multi-Class Output Codes},
  author={Maria-Florina Balcan and Travis Dick and Y. Mansour},
  booktitle={AAAI},
  year={2015}
}
  • Maria-Florina Balcan, Travis Dick, Y. Mansour
  • Published in AAAI 2015
  • Mathematics, Computer Science
  • We present a new perspective on the popular multi-class algorithmic techniques of one-vs-all and error correcting output codes. Rather than studying the behavior of these techniques for supervised learning, we establish a connection between the success of these methods and the existence of label-efficient learning procedures. We show that in both the realizable and agnostic cases, if output codes are successful at learning from labeled data, they implicitly assume structure on how the classes… CONTINUE READING

    Figures and Topics from this paper.

    Explore Further: Topics Discussed in This Paper

    References

    Publications referenced by this paper.
    SHOWING 1-10 OF 28 REFERENCES
    Combining labeled and unlabeled data with co-training
    4612
    A discriminative model for semi-supervised learning
    95
    Sensitive Error Correcting Output Codes
    82
    Multiclass Learning Approaches: A Theoretical Comparison with Implications
    32
    Co-Training and Expansion: Towards Bridging Theory and Practice
    257
    Reducing Multiclass to Binary: A Unifying Approach for Margin Classifiers
    1873
    Semi-Supervised Learning
    1876
    Theory of Disagreement-Based Active Learning
    116
    Two faces of active learning
    173