Daniël Vos

PhD candidate at Delft University of Technology


Room 4.E.420 / Building 28

Van Mourik Broekmanweg 6

2628 XE Delft

The Netherlands

Hi, I’m Daniël and I’m doing a PhD under the supervision of Dr. Sicco Verwer. The topic of my PhD is mainly on the safety and security of machine learning models and I have a special fondness for decision tree models. I particularly enjoy working on the discrete optimization problems that appear when training, certifying, and attacking robust decision trees. For example, with our method ROCT we turned the problem of training robust decision trees against adversarial examples into combinatorial optimization which can then be solved with MILP or MaxSAT solvers.

I strongly believe that we should be able to understand machine learning models that we deploy for critical applications and I worry that post-hoc XAI methods give us a false sense of understanding. Therefore I am also working on methods to train extremely small models with performance similar to more massive ones such as neural networks and gradient boosting ensembles. A small enough linear model or decision tree can be naturally understood by humans.

Next to my academic life, I love strength training, rowing, and field hockey. I also organize Delft’s hacking team, currently one of the highest-placed teams from the Netherlands on CTFTime.

selected publications

  1. IJCAI
    Optimal Decision Tree Policies for Markov Decision Processes
    Daniël Vos, and Sicco Verwer
    International Joint Conference on Artificial Intelligence 2023
  2. Adversarially Robust Decision Tree Relabeling
    Daniël Vos, and Sicco Verwer
    European Conference on Machine Learning and Principles and Practice of Knowledge Discovery in Databases 2022
  3. Robust optimal classification trees against adversarial examples
    Daniël Vos, and Sicco Verwer
    AAAI Conference on Artificial Intelligence 2022
  4. Efficient training of robust decision trees against adversarial examples
    Daniël Vos, and Sicco Verwer
    International Conference on Machine Learning 2021