I am a PhD Student supervised by Aad van der Vaart and Botond Szabo. I work in Bayesian nonparametric statistics, specifically in uncertainty quantification. However my interest is much broader than that, and I might take some side projects. Before my PhD, I did my bachelors and masters studies at Utrecht University and wrote my master thesis under the supervision of Aad van der Vaart.
My current projects:
- A Bernstein-Von Mises theorem for the Pitman-Yor process.
- Studying consistency, contraction rates and uncertainty quantification with a general class of proper species sampling priors, dubbed stick-breaking process priors, which have independent and identically distributed relative stick-breaking weights.
- General theory regarding structural and incidental parameters in a Bayesian model. For example the Neyman-Scott problem.
- Uncertainty quantification in deep learning.
News:
- I won the best Student/Postdoc Contributed Paper Award (ISBA 2021 world meeting)
- I gave a talk on the Bernstein-von Mises theorem for the Pitman-Yor processes on the ISBA
- I gave a talk on EcoSta on the frequentist coverage guarantees for empirical Bayesian Deep Neural networks.
- I gave two talks during the Bernoulli-IMS one world symposium 2020.
- I gave a talk during the cambridge zoom session on "Empirical Bayesian uncertainty quantification using deep neural networks in Besov spaces"
- I presented a poster at the BNP 2019 conference in Oxford on a Bernstein-Von Mises theorem for the Pitman-Yor process.
- I presented a poster at the conference for Aads Birthday in Leiden on a Bernstein-Von Mises theorem for the Pitman-Yor process.
- I gave a talk in The Bayes club on the Bernstein-Von Mises theorem for the Pitman-Yor process and uncertainty quantification with general stick-breaking process priors.
Research interests:
- Uncertainty quantification
- Deep learning
- Nonparametric regression
- Nonparametric causal inference
- Online learning