Anzeige
Erschienen in:
2019 | OriginalPaper | Buchkapitel
22. How to Prevent Bias in Self-Improving Models
verfasst von : Tobias Baer
Erschienen in: Understand, Manage, and Prevent Algorithmic Bias
Verlag: Apress
Aktivieren Sie unsere intelligente Suche, um passende Fachinhalte oder Patente zu finden.
Wählen Sie Textabschnitte aus um mit Künstlicher Intelligenz passenden Patente zu finden. powered by
Markieren Sie Textabschnitte, um KI-gestützt weitere passende Inhalte zu finden. powered by
Abstract
One of the greatest advantages of machine learning is that models can build and update themselves without any human intervention, enabling them to respond to structural changes at the fastest possible pace. The very context requiring such self-improving algorithms (the fast change of the environment in which they operate) is also the source of a heightened risk of biases affecting the algorithm, be it self-reinforcing feedback loops like we experienced in the context of social media (Chapter 11) or new data that might enable the algorithm to develop a bias against a protected class.