Corpus ID: 76663579

Learning Dependency Structures for Weak Supervision Models

@article{Varma2019LearningDS,
  title={Learning Dependency Structures for Weak Supervision Models},
  author={P. Varma and F. Sala and A. He and A. Ratner and C. R{\'e}},
  journal={ArXiv},
  year={2019},
  volume={abs/1903.05844}
}
  • P. Varma, F. Sala, +2 authors C. Ré
  • Published 2019
  • Computer Science, Mathematics
  • ArXiv
  • Labeling training data is a key bottleneck in the modern machine learning pipeline. [...] Key Method Under certain conditions, we show that the amount of unlabeled data needed can scale sublinearly or even logarithmically with the number of sources $m$, improving over previous efforts that ignore the sparsity pattern in the dependency structure and scale linearly in $m$. We provide an information-theoretic lower bound on the minimum sample complexity of the weak supervision setting. Our method outperforms weak…Expand Abstract
    Multi-Resolution Weak Supervision for Sequential Data
    7
    Train and You'll Miss It: Interactive Model Iteration with Weak Supervision and Pre-Trained Embeddings
    Weak Supervision with Incremental Source Accuracy Estimation
    • Richard Correro, rcorrero
    • 2019
    Weakly Supervised Sequence Tagging from Noisy Rules
    7
    Overton: A Data System for Monitoring and Improving Machine-Learned Products
    5
    Leveraging Multi-Source Weak Social Supervision for Early Detection of Fake News
    2
    Inspector Gadget: A Data Programming-based Labeling System for Industrial Images

    References

    Publications referenced by this paper.
    SHOWING 1-10 OF 51 REFERENCES
    Training Complex Models with Multi-Task Weak Supervision
    34
    Learning the Structure of Generative Models without Labeled Data
    67
    Inferring Generative Model Structure with Static Analysis
    27
    Data Programming: Creating Large Training Sets, Quickly
    258