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

Training datasets for machine learning often have some form of missingness.For example, to learn a model for deciding whom to give a loan, the availabletraining data includes individuals who were given a loan in the past, but notthose who were not. This missingness, if ignored, nullifies any fairnessguarantee of the training procedure when the model is deployed. Using causalgraphs, we characterize the missingness mechanisms in different real-worldscenarios. We show conditions under which various distributions, used inpopular fairness algorithms, can or can not be recovered from the trainingdata. Our theoretical results imply that many of these algorithms can notguarantee fairness in practice. Modeling missingness also helps to identifycorrect design principles for fair algorithms. For example, in multi-stagesettings where decisions are made in multiple screening rounds, we use ourframework to derive the minimal distributions required to design a fairalgorithm. Our proposed algorithm decentralizes the decision-making process andstill achieves similar performance to the optimal algorithm that requirescentralization and non-recoverable distributions.

The Importance of Modeling Data Missingness in Algorithmic Fairness: A Causal Perspective