Séminaire SCALab

Michaël Perrot, INRIA Lille

Fairness in Trustworthy Machine Learning.

The goal of this talk is to give an overview of my research activities on fairness in trustworthy machine learning and more precisely in classification. I will start the presentation with a high level introduction to fairness in machine learning, explaining where biases may come from, how they may be measured, and how they may be mitigated. Then, I will present some recent work where we study fairness as it interacts with other trustworthiness concepts.