jjzhu I am an associate professor in mathematics (tenured) at the KTH Royal Institute of Technology in Stockholm. Previously, I was the head of an independent research group at the Weierstrass Institute, Berlin. I am also a PI funded by DFG at TU Darmstadt. I worked as a postdoctoral researcher in machine learning at the Max Planck Institute for Intelligent Systems in Tübingen, Germany. My Ph.D. study was in optimization and numerical analysis, at the University of Florida. See here for a short bio. I also write a non-research blog here. However, the update frequency depends on how busy I am at the moment.

Overall, I am interested in computational algorithms and dynamical systems. My group focuses on the mathematical foundations of machine learning and optimization. In recent years, I have been interested in ML/OPT applications rooted in the principled theory of optimal transport, gradient flows, and kernel methods.

Specifically, I started my research career in optimization and subsequently became interested in robust machine learning and kernel methods. That requires us to use computational optimization tools that can manipulate probability distributions, which are inherently infinite-dimensional. It led me to my current interests in variational methods for machine learning and optimization over probability distributions, rooted in the theory of gradient flows and optimal transport.

For example, in some of my previous works, I invented robust probabilistic ML algorithms that can protect against distribution shifts using principled kernel methods. Those optimization algorithms have deep theoretical roots such as the analysis of PDEs. Following that and after moving to Berlin, I dedicate my current research to interfacing large-scale computational algorithms in machine learning/optimization using PDE gradient flows and optimal transport. Recently, I became interested in the Hellinger geometry (a.k.a. Fisher-Rao) and collaborated with Alexander Mielke on kernel methods and (Wasserstein-)Fisher-Rao, a.k.a. (spherical-)Hellinger-Kantorovich, gradient flows.

To get in touch, click the icon at the bottom of the page. There are sometimes delays in my response to emails, please be patient.

Upcoming events

Recent talks (selected)

Open positions

  • KTH Master thesis: if you are a master’s student already enrolled in KTH and interested in optimization for machine learning, deep generative models, optimal transport, applications of PDE/SDEs, please feel free to reach out.
  • [PhD position] If you are interested in joining my group at KTH, please feel free to inquire with a CV and all transcripts. I will read all inquires, but can only reply to those who fit with our group.
  • Joint PhD position at TU Darmstadt/KTH Royal Institute of Technology (with Jan Peters). See the ad here

News and updates

Twitter feed