From 10/2024, I will be Associate Professor at UCL’s Department of Statistical Science. Until 07/2025, I am fully bought out of teaching and administrative duties with an EPSRC research fellowship to continue my work on Optimisation-centric Generalisations of Bayesian Inference. I am also a visiting researcher at the Alan Turing Institute’s Data-Centric Engineering Programme, and scientific advisor for HopStair and Idoven. Additionally, I provide technical advice to the legal experts at FoxGlove to ensure that machine learning and AI are used for the common good, and work as a statistical expert witness in high-profile legal cases.
If you would like to learn more about my research, I have summarised some of my key contributions in the following talk. My research revolves around extending the paradigm of Bayesian inference to cope with the challenges posed by modern large-scale data, simulator models, and machine learning techniques. In this context, I am particularly interested in generalised and Post-Bayesian inference, model misspecification and robustification strategies, computational challenges involving intractability, and variational methods.
Prior to my current position, I was a Lecturer (07/2022-10/2024) and Biometrika Fellow in (10/2021-07/2022), also based at UCL. Before that, I was a doctoral candidate within the Oxford-Warwick Statistics programme (2016-2021) as well as the first UK-based Facebook Fellow (2020/2021). During that time, I also worked with the research arms of Amazon (2019) and DeepMind (2021). I remain eager to stay in close contact with industry, and am very open to being approached by potential industrial partners for both academic and non-academic work.
Download my (very likely outdated) resumé.
PhD in Statistical Sciences, 2022
OxWaSP (Oxford-Warwick Statistics Programme)
BSc+MSc in Econometrics & Operations Research, 2016
Maastricht University
My colleague Francois-Xavier Briol and I co-organise and co-run the machine learning research group at UCL’s department of statistical science. We organise a range of activities for and with our students, including joint research meetings and lunch every week, workshops, and social activities. Group members are actively encouraged to collaborate both within the group and beyond.
If you would like to work with me, it is advisable to make sure you are interested in the questions that drive my research. To learn more about them, I recommend you watch the following talk.
I have my own research group, and I take on new PhD students from diverse social and academic backgrounds. Candidates should have a strong background in statistics, computer science, mathematics, econometrics, physics, or a related discipline and be passionate about research on methodological or theoretical questions at the intersection between machine learning and Bayesian statistics. It is important to me that my students are supervised well. Since I have no teaching or administrative duties over the coming years, I can focus on supervision for a large part of my time.
If you are interested, please do get in touch with me directly and send through
You are not required to have a clear thought-out research plan before getting in touch with me (though it is encouraged). Please email me at j dot knoblauch at ucl dot ac dot uk.
In case you are going to apply with me, I also recommend that you reach out to my current/former students and ask them what it is like (see list below): get a first-hand outlook on what it is like to be supervised by me to help you decide if I would be a good fit for you. I recommend you do this with any supervisor you are considering to get a clearer picture of the kind of 3-4 year commitment you are entering into.
I am offering a few master’s projects this year that will be circulated at the department. If you are interested in any of these or have an alternative idea that you think would fit within my research remit, please get in touch with me by sending me an email.