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 the United States and elsewhere, risk assessment algorithms are being usedto help inform criminal justice decision-makers. A common intent is to forecastan offender's ``future dangerousness.'' Such algorithms have been correctlycriticized for potential unfairness, and there is an active cottage industrytrying to make repairs. In this paper, we use counterfactual reasoning toconsider the prospects for improved fairness when members of a less privilegedgroup are treated by a risk algorithm as if they are members of a moreprivileged group. We combine a machine learning classifier trained in a novelmanner with an optimal transport adjustment for the relevant joint probabilitydistributions, which together provide a constructive response to claims ofbias-in-bias-out. A key distinction is between fairness claims that areempirically testable and fairness claims that are not. We then use confusiontables and conformal prediction sets to evaluate achieved fairness forprojected risk. Our data are a random sample of 300,000 offenders at theirarraignments for a large metropolitan area in the United States during whichdecisions to release or detain are made. We show that substantial improvementin fairness can be achieved consistent with a Pareto improvement for protectedgroups.

Improving Fairness in Criminal Justice Algorithmic Risk Assessments Using Optimal Transport and Conformal Prediction Sets