We gratefully acknowledge support from
the Simons Foundation and member institutions.
Full-text links:

Download:

Current browse context:

cs.LG

Change to browse by:

References & Citations

DBLP - CS Bibliography

Bookmark

(what is this?)
CiteULike logo BibSonomy logo Mendeley logo del.icio.us logo Digg logo Reddit logo ScienceWISE logo

Computer Science > Machine Learning

Title: End-to-End Weak Supervision

Abstract: Aggregating multiple sources of weak supervision (WS) can ease the data-labeling bottleneck prevalent in many machine learning applications, by replacing the tedious manual collection of ground truth labels. Current state of the art approaches that do not use any labeled training data, however, require two separate modeling steps: Learning a probabilistic latent variable model based on the WS sources -- making assumptions that rarely hold in practice -- followed by downstream model training. Importantly, the first step of modeling does not consider the performance of the downstream model. To address these caveats we propose an end-to-end approach for directly learning the downstream model by maximizing its agreement with probabilistic labels generated by reparameterizing previous probabilistic posteriors with a neural network. Our results show improved performance over prior work in terms of end model performance on downstream test sets, as well as in terms of improved robustness to dependencies among weak supervision sources.
Comments: Code URL: this https URL
Subjects: Machine Learning (cs.LG); Artificial Intelligence (cs.AI); Machine Learning (stat.ML)
Journal reference: Thirty-fifth Conference on Neural Information Processing Systems (NeurIPS 2021)
Cite as: arXiv:2107.02233 [cs.LG]
  (or arXiv:2107.02233v3 [cs.LG] for this version)

Submission history

From: Salva Rühling Cachay [view email]
[v1] Mon, 5 Jul 2021 19:10:11 GMT (7040kb,D)
[v2] Sat, 30 Oct 2021 11:35:50 GMT (4323kb,D)
[v3] Tue, 30 Nov 2021 05:09:04 GMT (4324kb,D)

Link back to: arXiv, form interface, contact.