About me

I am an associate professor in the Seminar for Statistics of ETH D-MATH. Previously, I was an assistant professor in the Department of Statistical Science at Duke University from Spring 2021 to Spring 2024. I was a postdoc fellow at ETH Foundations of Data Science (ETH-FDS) in ETH Zürich under the supervision of Prof. Peter Bühlmann. I obtained my PhD in the Department of Statistics at UC Berkeley in 2019. I was very fortunate to be advised by Prof. Bin Yu. During my PhD, I was also fortunate to work with Prof. Martin Wainwright and Prof. Jack Gallant. Before my PhD study, I obtained my Diplome d'Ingénieur (Eng. Deg. in Applied Mathematics) at Ecole Polytechnique in France.

My main research interests lie on statistical machine learning, MCMC sampling, high dimension geometry/concentration of measure, domain adaptation and statistical challenges that arise in computational neuroscience.

Q: Why does a Markov chain sampling researcher care about high-dimensional geometry/concentration of measure?
A: If sampling algorithms are Formula 1 cars, then concentration of measure is the geometry of the race track they run on.

Github
Google Scholar

In the news: Quanta Magazine article about the first almost-constant bound on Kannan-Lovász-Simonovits (KLS) Conjecture and Bourgain's slicing problem.

Current group members

• Minhui Jiang (PhD student)
• Almut Rödder (PhD student, co-supervised with Afonso Bandeira)
• Matthieu Dages (PhD student)
• Peter Whalley (postdoc)
• Francesco Pedrotti (postdoc)

For prospective students

• If you already have a master degree or are about to receive one, apply directly to the Zurich Graduate School of Mathematics for a position in the Department of Mathematics (two deadlines a year)
• Otherwise, you may consider applying to master programmes in D-MATH

For current ETHZ undergraduate students and master students looking for a undergraduate/master thesis topic

You may take a look at the following topics on my Google Scholar and see if there is a fit

• Markov chain Monte Carlo sampling algorithms, theory and coding. In terms of past projects,
  • for theory, take a look at our paper on the mixing of Metropolis-Adjusted Langevin Algorithm.
  • for code, take a look at our latest open-source implementation of PolytopeWalk on Github. We welcome developers of all level.
• Statistical learning under distributional shifts (or domain adaptation, empirical Bayes)
• Statistical modeling in computational neuroscience: calcium imaging data from mice visual cortex