Make “Fairness by Design” Part of Machine Learning
Harvard Business
AUGUST 1, 2018
More recently, a study compared the commonly used crime risk analysis tool COMPAS against recidivism predictions from 400 untrained workers recruited via Amazon Mechanical Turk. Sampling bias may produce models trained on data that is not fully representative of future cases. Pair data scientists with a social scientist.
30 
Let's personalize your content