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

In recent years fairness in machine learning (ML) has emerged as a highlyactive area of research and development. Most define fairness in simple terms,where fairness means reducing gaps in performance or outcomes betweendemographic groups while preserving as much of the accuracy of the originalsystem as possible. This oversimplification of equality through fairnessmeasures is troubling. Many current fairness measures suffer from both fairnessand performance degradation, or "levelling down," where fairness is achieved bymaking every group worse off, or by bringing better performing groups down tothe level of the worst off. When fairness can only be achieved by makingeveryone worse off in material or relational terms through injuries of stigma,loss of solidarity, unequal concern, and missed opportunities for substantiveequality, something would appear to have gone wrong in translating the vagueconcept of 'fairness' into practice. This paper examines the causes andprevalence of levelling down across fairML, and explore possible justificationsand criticisms based on philosophical and legal theories of equality anddistributive justice, as well as equality law jurisprudence. We find thatfairML does not currently engage in the type of measurement, reporting, oranalysis necessary to justify levelling down in practice. We propose a firststep towards substantive equality in fairML: "levelling up" systems by designthrough enforcement of minimum acceptable harm thresholds, or "minimum rateconstraints," as fairness constraints. We likewise propose an alternativeharms-based framework to counter the oversimplified egalitarian framingcurrently dominant in the field and push future discussion more towardssubstantive equality opportunities and away from strict egalitarianism bydefault. N.B. Shortened abstract, see paper for full abstract.

The Unfairness of Fair Machine Learning: Levelling down and strict egalitarianism by default