David Corliss examines how shifting the focus to testing for and mitigating bias at the design stage instead of after code is released can help prevent many of the problems seen in machine learning and artificial intelligence.
Amstat News
The Membership Magazine of the American Statistical Association