End-to-End Weak Supervision

Salva Rühling Cachay, Benedikt Boecking, Artur Dubrawski

Keywords: weak supervision, deep learning, data programming

TL;DR: A neural end-to-end system that learns exclusively from multiple sources of 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.

Code Of Conduct: I certify that all co-authors of this work have read and commit to adhering to the NeurIPS Statement on Ethics, Fairness, Inclusivity, and Code of Conduct.

Code: https://github.com/autonlab/weasel

Community Implementations: [ 2 code implementations](https://www.catalyzex.com/paper/arxiv:2107.02233/code)