Fairlearn: A Toolkit for Assessing and Improving Fairness in AI

Focus: Methods or Design

Source: Microsoft

Readability: Intermediate

Type: PDF Article

Open Source: Yes

External URL: https://www.microsoft.com/en-us/research/uploads/prod/2020/05/Fairlearn_whitepaper.pdf

Keywords: N/A

Learn Tags: Bias Data Tools Design/Methods Fairness Framework Inclusive Practice Solution Trust

Summary: Fairlearn is an open source toolkit that empowers data scientists and developers to assess and improve the fairness of their AI systems through an interactive visualization dashboard and unfairness mitigation algorithms.