Consistency versus Realizable H-Consistency for Multiclass Classification

@inproceedings{Long2013ConsistencyVR,
  title={Consistency versus Realizable H-Consistency for Multiclass Classification},
  author={Philip M. Long and Rocco A. Servedio},
  booktitle={ICML},
  year={2013}
}
A consistent loss function for multiclass classification is one such that for any source of labeled examples, any tuple of scoring functions that minimizes the expected loss will have classification accuracy close to that of the Bayes optimal classifier. While consistency has been proposed as a desirable property for multiclass loss functions, we give experimental and theoretical results exhibiting a sequence of linearly separable data sources with the following property: a multiclass… CONTINUE READING

Citations

Publications citing this paper.

References

Publications referenced by this paper.
SHOWING 1-10 OF 19 REFERENCES

Statistical analysis of some multi-category large margin classification methods

T. Zhang
  • Journal of Machine Learning Research,
  • 2004
VIEW 10 EXCERPTS
HIGHLY INFLUENTIAL

Large-scale image classification with trace-norm regularization

  • 2012 IEEE Conference on Computer Vision and Pattern Recognition
  • 2012
VIEW 2 EXCERPTS

Towards good practice in large-scale learning for image classification

  • 2012 IEEE Conference on Computer Vision and Pattern Recognition
  • 2012
VIEW 1 EXCERPT