AI: The issue of bias
Vincent Violago and Nikko Quevada discuss the various types of bias in machine learning algorithms and the data used to train the algorithms. They also analyse patents directed to bias mitigation, as well as the reasons machine bias is difficult to identify and eliminate
Machine or algorithmic biasAlong with artificial
intelligence (AI) interpretability and reproducibility issues,
debiasing or the elimination of bias relating to the
application of AI particularly in healthcare, criminal justice,
education and welfare has become one of the AI field's
Please log in
to read the rest of this article. New to Managing Intellectual Property?
Take advantage of free access to up to 5 articles on Managing IP and become a member today. It’s free to join and the benefits start straight away.
Please make sure you log in to read the rest of the article.
Join us nowGain FREE access to up to five free articles when you register now.