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

Machine learning (ML) is playing an increasingly important role in renderingdecisions that affect a broad range of groups in society. ML models informdecisions in criminal justice, the extension of credit in banking, and thehiring practices of corporations. This posits the requirement of modelfairness, which holds that automated decisions should be equitable with respectto protected features (e.g., gender, race, or age) that are oftenunder-represented in the data. We postulate that this problem ofunder-representation has a corollary to the problem of imbalanced datalearning. This class imbalance is often reflected in both classes and protectedfeatures. For example, one class (those receiving credit) may beover-represented with respect to another class (those not receiving credit) anda particular group (females) may be under-represented with respect to anothergroup (males). A key element in achieving algorithmic fairness with respect toprotected groups is the simultaneous reduction of class and protected groupimbalance in the underlying training data, which facilitates increases in bothmodel accuracy and fairness. We discuss the importance of bridging imbalancedlearning and group fairness by showing how key concepts in these fields overlapand complement each other; and propose a novel oversampling algorithm, FairOversampling, that addresses both skewed class distributions and protectedfeatures. Our method: (i) can be used as an efficient pre-processing algorithmfor standard ML algorithms to jointly address imbalance and group equity; and(ii) can be combined with fairness-aware learning algorithms to improve theirrobustness to varying levels of class imbalance. Additionally, we take a steptoward bridging the gap between fairness and imbalanced learning with a newmetric, Fair Utility, that combines balanced accuracy with fairness.

Towards A Holistic View of Bias in Machine Learning: Bridging Algorithmic Fairness and Imbalanced Learning