Rianne de Heide
About me
I am an assistant professor in the Department of Mathematics, at the Vrije Universiteit Amsterdam. I work on problems and solutions in machine learning and statistics. My research focuses on sequential learning, and in particular on (sequential) hypothesis testing, Bayesian learning and best-arm identification problems. Central in my work is both bringing a solid mathematical foundation to the topics I work on in different fields, as well as making the theory accessible for less mathematical audiences. I’m also interested in the mathematical and philosophical foundations of Bayesianism, machine learning, statistics and probability theory.
Find out more.
PhD position
I have an opening for a fully funded 4-year PhD position on Mathematical Statistics and Machine Learning Theory. If you want to work on e-values, multiple testing, bandits, Bayesian inference, or combine all of those, please apply before August 1st. Feel free to e-mail me if you have any questions (but apply via the website below).
link to the vacancy
We are an inclusive group and diversity is at the heart of our research principles. We care about a good working atmosphere and a good work-life balance. Applications from all groups currently under-represented in academic posts are especially encouraged.
News
June 25: I give a talk at Zorginstituut Nederland about e-values, - I think this is my first e-value talk in Dutch!
June 25: I give a talk at the Epidemiology group of the LUMC (website) about e-values.
June 24: I give a talk at RIVM (Rijksinstituut voor Volksgezondheid en Milieu) about e-values.
In September I will move to the University of Twente.
Our paper on GROW-optimal group-invariant statistics (arxiv) is accepted to the Annals of Statistics.
May 5-9: I had a fantastic week with fellow e-value enthusiasts in Oberwolfach.
April 26: Together with Jelle Goeman we submitted a grant proposal (M2) to NWO.
There were several interviews with me the last months in (national) newspapers and magazines: NAW (PDF), NRC (PDF, Link), The New Scientist (Picture), and Ad Valvas (Link).
Older News
More old news
Research
My research interests include
Bandits, Reinforcement Learning
(Sequential) hypothesis testing, multiple testing
Group invariance in statistics
Bayesian methods
Learning theory
Imprecise probabilities in statistics
Foundations of ML, stats and probability theory
Find out more.
Recent Publications
Safe Testing
Peter Grünwald, Rianne de Heide and Wouter Koolen
Forthcoming as disscussion paper at the JRSS-B, 2024 arxiv preprint RSS
The discussion meeting took place on January 24, 2024.
Attribution-based Explanations that Provide Recourse Cannot be Robust
Hidde Fokkema, Rianne de Heide and Tim van Erven
arXiv 2205.15834, 2023 arxiv proc
Journal of Machine Learning Research, 24(360), pp.1-37.
|