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/

Zendy Plus

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

Zendy Tools

Zendy Open

As machine learning is increasingly used to make real-world decisions, recentresearch efforts aim to define and ensure fairness in algorithmic decisionmaking. Existing methods often assume a fixed set of observable features todefine individuals, but lack a discussion of certain features not beingobserved at test time. In this paper, we study fairness of naive Bayesclassifiers, which allow partial observations. In particular, we introduce thenotion of a discrimination pattern, which refers to an individual receivingdifferent classifications depending on whether some sensitive attributes wereobserved. Then a model is considered fair if it has no such pattern. We proposean algorithm to discover and mine for discrimination patterns in a naive Bayesclassifier, and show how to learn maximum likelihood parameters subject tothese fairness constraints. Our approach iteratively discovers and eliminatesdiscrimination patterns until a fair model is learned. An empirical evaluationon three real-world datasets demonstrates that we can remove exponentially manydiscrimination patterns by only adding a small fraction of them as constraints.

Learning Fair Naive Bayes Classifiers by Discovering and Eliminating Discrimination Patterns