Author: d’Alessandro B, O’Neil C, LaGatta T.
Publisher: National Library of Medicine
Publication Year: 2017
Summary: The following article describes discrimination in the context of machine learning. The authors discuss how models can be biased even when the researcher has the best intentions. The authors illustrate a process in which the data scientist can be discrimination aware, injecting discrimination awareness into the standard data science workflow. This workflow includes discrimination aware testing during the exploratory data analysis phase. The authors walk the reader though case studies for illustration