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

Existing algorithms for ensuring fairness in AI use a single-shot trainingstrategy, where an AI model is trained on an annotated training dataset withsensitive attributes and then fielded for utilization. This training strategyis effective in problems with stationary distributions, where both training andtesting data are drawn from the same distribution. However, it is vulnerablewith respect to distributional shifts in the input space that may occur afterthe initial training phase. As a result, the time-dependent nature of data canintroduce biases into the model predictions. Model retraining from scratchusing a new annotated dataset is a naive solution that is expensive andtime-consuming. We develop an algorithm to adapt a fair model to remain fairunder domain shift using solely new unannotated data points. We recast thislearning setting as an unsupervised domain adaptation problem. Our algorithm isbased on updating the model such that the internal representation of dataremains unbiased despite distributional shifts in the input space. We provideextensive empirical validation on three widely employed fairness datasets todemonstrate the effectiveness of our algorithm.

Preserving Fairness in AI under Domain Shift