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 Tools annual

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 monthly

Global: Zendy Plus monthly

Presents the access of premium content as premium feature

Premium Content

Global: Zendy Plus annual

Global: Zendy Free Plan

Dataset bias is one of the prevailing causes of unfairness in machinelearning. Addressing fairness at the data collection and dataset preparationstages therefore becomes an essential part of training fairer algorithms. Inparticular, active learning (AL) algorithms show promise for the task bydrawing importance to the most informative training samples. However, theeffect and interaction between existing AL algorithms and algorithmic fairnessremain under-explored. In this paper, we study whether models trained withuncertainty-based AL heuristics such as BALD are fairer in their decisions withrespect to a protected class than those trained with identically independentlydistributed (i.i.d.) sampling. We found a significant improvement on predictiveparity when using BALD, while also improving accuracy compared to i.i.d.sampling. We also explore the interaction of algorithmic fairness methods suchas gradient reversal (GRAD) and BALD. We found that, while addressing differentfairness issues, their interaction further improves the results on mostbenchmarks and metrics we explored.

Can Active Learning Preemptively Mitigate Fairness Issues?