Bayesian Weak Supervision via an Optimal Transport Approach

Putra Manggala, Holger Hoos, Eric Nalisnick

Keywords: weak supervision, probabilistic modelling, optimal transport

TL;DR: We propose a Bayesian probabilistic model that employs a tractable Sinkhorn-based optimal transport formulation to derive a ground-truth label.

Abstract: Large-scale machine learning is often impeded by a lack of labeled training data. To address this problem, the paradigm of weak supervision aims to collect and then aggregate multiple noisy labels. We propose a Bayesian probabilistic model that employs a tractable Sinkhorn-based optimal transport formulation to derive a ground-truth label. The translation between true and weak labels is cast as a transport problem with an inferred cost structure. Our approach achieves strong performance on the WRENCH weak supervision benchmark. Moreover, the posterior distribution over cost matrices allows for exploratory analysis of the weak sources.