About us
The Fundamentals of Statistical Machine Learning group is a research group in the UCL Department of Statistical Science. Our focus is on the intersection of statistical inference and machine learning methodology and theory.
Meet the Team
Faculty
Alessandro Barp
Assistant Professor
François-Xavier Briol
Associate Professor
Jeremias Knoblauch
Assistant Professor
EPSRC Fellow
Postdoctoral Researchers
Ieva Kazlauskaitė
IMSS Fellow
PhD Students
Matias Altamirano
PhD Student
Zonghao (Hudson) Chen
PhD Student
Oscar Key
PhD Student
Yann McLatchie
PhD Student
Masha Naslidnyk
PhD Student
Iman Simo-Dzumgang
PhD Student
Ilina Yozova
PhD Student
Visiting Researchers
Xing Liu
PhD Student
Saifuddin Syed
Florence Nightingale Bicentennial Fellow
Alumni
Ayush Bharti
Visitor, Autumn 2022
Kaiyu Li
PhD, 2019-2024
Zhuo Sun
PhD, 2019-2023
Veit Wild
Visitor, Autumn 2023
News
Our paper Robust and Conjugate Gaussian Process Regression was accepted as a spotlight paper (top 3.5%) at ICML 2024!
One paper accepted at UAI 2024 on the topic of Bayesian quadrature for computing parametric expectations.
Two papers accepted at ICML 2024. The papers are both on the topic of scalable generalised Bayes, one in the context of Gaussian processes, and the other in the context of Kalman filtering.
Many members of the group attend and present at the ‘Functional inference and machine intelligence’ workshop at the University of Bristol.
Iman Simo-Dzumgang joins the group as a PhD student co-supervised by Dr. F-X Briol and Dr. Jeremias Knoblauch.
Recent Publications
(2024).
Conditional Bayesian Quadrature.
In UAI.
(2024).
Outlier-robust Kalman Filtering through Generalised Bayes.
In ICML.
(2024).
Robust and Conjugate Gaussian Process Regression.
In ICML.
(2024).
The Impact of Loss Estimation on Gibbs Measures.
In arXiv.
(2024).
A Rigorous Link between Deep Ensembles and (Variational) Bayesian Methods.
In NeurIPS.
(2023).
Multilevel Bayesian Quadrature.
In AISTATS.
(2023).
No Train No Gain: Revisiting Efficient Training Algorithms For Transformer-based Language Models.
In NeurIPS.
(2023).
Robustifying likelihoods by optimistically re-weighting data.
In arXiv.
(2023).
Vector-Valued Control Variates.
In ICML.
Major Research Funding
EPSRC New Investigator Award. Project on ‘Transfer Learning for Monte Carlo Methods’ (PI: Briol)
EPSRC Small Grant in the Mathematical Sciences, Project on “Robust Foundations for Bayesian Inference” (PI: Briol, co-I: Knoblauch)
UKRI/Turing Project “Digital Twin Handbook” (co-I: Briol)
EP/W005859/1: EPSRC Fellowship, Project on “Optimisation-centric Generalisations of Bayesian Inference” (PI: Knoblauch)
Biometrika fellowship: “Generalising Bayesian Inference” (PI: Knoblauch)
Amazon Research Award: Project on “Transfer Learning for Numerical Integration in Expensive Machine Learning Systems” (PI: Briol)
Contact
- 1-19 torrington place, London, WC1E 7HB