Stochastic Negative Mining for Learning with Large Output Spaces

@article{Reddi2018StochasticNM,
  title={Stochastic Negative Mining for Learning with Large Output Spaces},
  author={Sashank J. Reddi and Satyen Kale and Felix X. Yu and Daniel N. Holtmann-Rice and Jiecao Chen and Sanjiv Kumar},
  journal={ArXiv},
  year={2018},
  volume={abs/1810.07076}
}
We consider the problem of retrieving the most relevant labels for a given input when the size of the output space is very large. Retrieval methods are modeled as set-valued classifiers which output a small set of classes for each input, and a mistake is made if the label is not in the output set. Despite its practical importance, a statistically principled, yet practical solution to this problem is largely missing. To this end, we first define a family of surrogate losses and show that they… CONTINUE READING
1
Twitter Mention

Similar Papers

Tables and Topics from this paper.

References

Publications referenced by this paper.
SHOWING 1-10 OF 34 REFERENCES

Classification with a Reject Option using a Hinge Loss

  • J. Mach. Learn. Res.
  • 2008
VIEW 5 EXCERPTS
HIGHLY INFLUENTIAL