PROPOSAL

Predicting popularity of machine learning challenges


Supervisors: Veronika Cheplygina
Semester: Fall 2021
Tags: machine learning, data science, medical imaging

Machine learning challenges hosted on platforms such as Kaggle (general) or grand-challenge.org (medical imaging) have attracted a lot of attention, both from academia and industry researchers. Challenge designs vary widely [1], including what type of data is available, how the algorithms are evaluated, and the rewards for the winners. In medical imaging, there is some evidence that challenges might be creating a shift in attention to different diseases [2], for example, there is a disproportionate increase in papers on lung cancer after the 2016 Kaggle challenge on the subject. The goal of the project is to investigate trends in challenge participation, and/or their effects on the research field after the challenge. The exact research questions will be decided together with the student or students (group projects possible).

[1] Maier-Hein, Lena, et al. “Why rankings of biomedical image analysis competitions should be interpreted with care.” Nature communications 9.1 (2018): 1-13.

[2] Varoquaux, GaĆ«l, and Veronika Cheplygina. “How I failed machine learning in medical imaging–shortcomings and recommendations.” arXiv preprint arXiv:2103.10292 (2021).