What it is about:
With machine learning algorithms being applied to more and more decisions that directly affect people, we need to ask how we can translate concepts of fairness and non-discrimination into formulations that machines can understand and optimize.
What we work on:
In this topic group we review current research on “Fairness in Machine Learning” and discuss what different “fairness criteria” can and cannot achieve. We plan to share our perspectives in a blog post, with a visualization explaining the criteria, that is accessible to non-experts. Furthermore, we plan and develop workshops for various events. We aim to design the workshops for different audiences with different levels of knowledge: from experts to layman.
Where to find our results:
- Workshop at Kösk
- Workshop/Lecture for the course “Maschinelle Intelligenz und Gesellschaft” at TUM
- Blog article on “fairness criteria”
Who we are:
Chiara, Cathy, Julia, Teodora, Gunnar, Julius