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 Tools annual

Global: Zendy Free Plan

Global: Zendy Tools monthly

Global: Zendy Plus monthly

In real world datasets, particular groups are under-represented, much rarerthan others, and machine learning classifiers will often preform worse onunder-represented populations. This problem is aggravated across many domainswhere datasets are class imbalanced, with a minority class far rarer than themajority class. Naive approaches to handle under-representation and classimbalance include training sub-population specific classifiers that handleclass imbalance or training a global classifier that overlooks sub-populationdisparities and aims to achieve high overall accuracy by handling classimbalance. In this study, we find that these approaches are vulnerable in classimbalanced datasets with minority sub-populations. We introduced Fair-Net, abranched multitask neural network architecture that improves bothclassification accuracy and probability calibration across identifiablesub-populations in class imbalanced datasets. Fair-Nets is a straightforwardextension to the output layer and error function of a network, so can beincorporated in far more complex architectures. Empirical studies with threereal world benchmark datasets demonstrate that Fair-Net improves classificationand calibration performance, substantially reducing performance disparitybetween gender and racial sub-populations.

Fair-Net: A Network Architecture For Reducing Performance Disparity Between Identifiable Sub-Populations