Data Ethics and Protection: Law, Rights and Responsible Practice

Algorithms, AI and automated decisions

Bias in, bias out

How bias enters a model through biased training data and proxy variables, shown with two real cases: Amazon's scrapped hiring tool and the disputed COMPAS criminal risk scores.

This lesson is part of the paid path

Enrol in Data Ethics and Protection: Law, Rights and Responsible Practice to unlock this lesson and the rest of the course. The preview lessons stay free.

$79 $49
Sign in to enrol Prefer to see more first? Back to the course and read the free previews.