English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Plus annual

Presents the access of premium content as premium feature

Premium Content

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Plus monthly

Global: Zendy Tools annual

Global: Zendy Tools monthly

Global: Zendy Free Plan

In weakly supervised learning, unbiased risk estimator(URE) is a powerfultool for training classifiers when training and test data are drawn fromdifferent distributions. Nevertheless, UREs lead to overfitting in many problemsettings when the models are complex like deep networks. In this paper, weinvestigate reasons for such overfitting by studying a weakly supervisedproblem called learning with complementary labels. We argue the quality ofgradient estimation matters more in risk minimization. Theoretically, we showthat a URE gives an unbiased gradient estimator(UGE). Practically, however,UGEs may suffer from huge variance, which causes empirical gradients to beusually far away from true gradients during minimization. To this end, wepropose a novel surrogate complementary loss(SCL) framework that trades zerobias with reduced variance and makes empirical gradients more aligned with truegradients in the direction. Thanks to this characteristic, SCL successfullymitigates the overfitting issue and improves URE-based methods.

Unbiased Risk Estimators Can Mislead: A Case Study of Learning with Complementary Labels