AI Bias Mitigation: A Developer’s Guide
Detection and mitigation of bias in Artificial Intelligence (AI) projects isn’t just a technical nuisance. It’s a systemic design challenge, and fixing it requires collaboration between engineers, product owners, and leadership.
Praktische info:
Inschrijven?
- Voorwaarden: Familiarity with python and basic machine learning concepts
- Prijs: Free of charge for all participants from academia, industry, and public administration from EU and/or EuroHPC JU member countries
georganiseerd door:
Leertraject
Building on our previous webinar Machine Learning and Human Prejudice: Understanding Bias in AI, this hands-on workshop brings the conversation into the codebase. Participation in the webinar is not a prerequisite for this course.
This session is designed for those who build, train, and deploy machine learning models, and who want to move beyond awareness into concrete action. Through guided notebook walkthroughs and a hands-on exercises, participants will explore how bias enters AI systems at the data, algorithm, and interpretation level, learn to measure fairness using industry-standard metrics, and apply mitigation techniques using libraries like IBM’s AI Fairness 360. The session also covers documentation practices and post-deployment monitoring, so that fairness becomes a built-in habit rather than an afterthought.
Gerelateerde opleidingen
