Spectral Methods meet EM: A Provably Optimal Algorithm for Crowdsourcing

@article{Zhang2014SpectralMM,
  title={Spectral Methods meet EM: A Provably Optimal Algorithm for Crowdsourcing},
  author={Yuchen Zhang and Xi Chen and Dengyong Zhou and Michael I. Jordan},
  journal={Journal of Machine Learning Research},
  year={2014},
  volume={17},
  pages={102:1-102:44}
}
Crowdsourcing is a popular paradigm for effectively collecting labels at low cost. The DawidSkene estimator has been widely used for inferring the true labels from the noisy labels provided by non-expert crowdsourcing workers. However, since the estimator maximizes a non-convex loglikelihood function, it is hard to theoretically justify its performance. In this paper, we propose a two-stage efficient algorithm for multi-class crowd labeling problems. The first stage uses the spectral method to… CONTINUE READING

From This Paper

Figures, tables, and topics from this paper.

Citations

Publications citing this paper.
SHOWING 1-10 OF 139 CITATIONS, ESTIMATED 36% COVERAGE

FILTER CITATIONS BY YEAR

2013
2019

CITATION STATISTICS

  • 34 Highly Influenced Citations

  • Averaged 38 Citations per year over the last 3 years

  • 5% Increase in citations per year in 2018 over 2017

References

Publications referenced by this paper.

Similar Papers

Loading similar papers…