About Me
Hi there! my name is Floris, and I am a PhD candidate in Machine Learning at the University of Amsterdam.
My ambition is to make a contribution towards the responsible and fair use of machine learning algorithms in society. Currently, I am pursuing this ambition by developing methods to prevent deep neural networks from relying on spurious correlations.
In addition, I am a researcher at Algorithm Audit, helping public organisations with the ethical use of algorithms.
In a past life, I was ranked as one of the top 10 university debaters in Europe, and spoke in the finals of the World University Debate Championships. I also have industry experience at the largest hedge fund in the world (Bridgewater).
Click here for my CV.
Research
Optimizing importance weighting in the presence of sub-population shifts
International Conference on Learning Representations (ICLR), 2025
Removing Spurious Concepts from Neural Network Representations via Joint Subspace Estimation
International Conference of Machine Learning (ICML), 2024
Policy
Next to my PhD, I try to contribute to making algorithms fairer and responsible through collaborating with policy makers and open-source software. This work is primarily done as part of Algorithm Audit.
Developing an open-source unsupervised bias detection tool
Currently working on a tool that can be used by public organisations to identify potentially unfairly treated groups of similar users. The tool can be easily used here, and our repository can be found here.
External investigation into an algorithm of the Dutch Education Executive Agency
Led the quantitative research for Algorithm Audit into an algorithm used by Dutch Education Executive Agency (DUO) that was suspected to be profiling against students of non-European migration backgrounds. Based on conclusions from our initial report, the Minister of education apologised for the use of the Algorithm. In a second report, after obtaining data from the Central Bureau of Statistics (CBS), we investigated the sources of predjudice for the algorithm. The second report can be found here here, and our code to reproduce findings from the report here. A working paper with our findings can be found here
Teaching
Supervising Bachelor & Master Theses
Supervised several students over the course of two months., University of Amsterdam, Bachelor of Econometrics and Data Science,
Teaching coursework in Statistical Learning and Reinforcement Learning
2nd and 3rd year courses, University of Amsterdam, Bachelor of Econometrics and Data Science,