Fairness and Inclusivity in Machine Learning Competitions

Machine learning competitions are increasingly used to develop, evaluate, and compare machine learning methods. They bring together participants from different countries, institutions, disciplines, and sectors, including universities, companies, and independent communities.

However, despite their growing importance, we know relatively little about who actually participates in these competitions. While competitions may be open to broad participation in principle, participation may vary depending on factors such as the type of platform, required expertise, access to data and computational resources, or institutional support. The goal of this project is:

• To investigate participation in medical imaging competitions by systematically collecting and analysing information about participants across a selected set of competitions. This may include information about participating teams, their geographical and institutional affiliations, organisational sectors, team structures, and other characteristics where available.
• It may also examine the extent to which observed patterns of participation align with, or raise questions about, broader expectations of representation, inclusion, and fairness, for example in relation to geographical, institutional, gender, or other forms of representation where data are available.

Groups of 2+ students (from any study program, mixed groups are welcome).

References

Eisenmann, M., Reinke, A., Weru, V., Tizabi, M. D., Isensee, F., Adler, T. J., … & Filipiak, P. (2022). Biomedical image analysis competitions: The state of current participation practice. arXiv preprint arXiv:2212.08568