Statistical Machine Learning Group

Research group

University College London

We are a research group at UCL’s Centre for Artificial Intelligence. Our research expertise is in data-efficient machine learning, probabilistic modeling, and autonomous decision making. Applications focus on robotics, climate science, and sustainable development.

If you are interested in joining the team, please check out our openings.

Meet the Team

Principal Investigators

Avatar

Marc Deisenroth

DeepMind Chair of Machine Learning and Artificial Intelligence

Machine learning, Gaussian processes, Reinforcement learning, Robotics, Meta learning

Research Fellows

Avatar

Yasemin Bekiroğlu

Senior Research Fellow

Machine learning, Robotics

Avatar

So Takao

Research Fellow

Machine learning, Climate science, Fluid mechanics, Geometric mechanics

PhD Students

Avatar

Alexander Terenin

PhD Student

Machine learning, Bayesian theory, Differential-geometric learning, Structural priors

Avatar

Steindór Sæmundsson

PhD Student

Machine learning, Gaussian processes, Meta learning, Structural priors, Variational inference

Avatar

Janith Petangoda

PhD Student

Machine learning, Meta learning, Differential geometry, Reinforcement learning

Avatar

Samuel Cohen

PhD Student

Machine learning, Optimal transport, Gaussian processes

Avatar

James Wilson

PhD Student

Machine learning, Gaussian processes, Bayesian optimization, Practical approximate inference

Avatar

Yicheng Luo

PhD Student

Robotics, Meta learning, Probabilistic programming

Avatar

Jackie Kay

PhD Student

Machine learning, Robotics, Reinforcement Learning, Meta learning

Avatar

Jacob Menick

PhD Student

Machine learning, Generative models, Large-scale deep learning, Variational inference, Information theory, Sparsity

Avatar

Mihaela Rosca

PhD Student

Generative models, Reinforcement learning, Natural language processing, Scalable and safe machine learning.

Avatar

Sicelukwanda Zwane

PhD Student

Machine learning, Robotics, Transfer Learning, Reinforcement Learning

Undergraduates

Avatar

Antoine Khoury

MSc Student

Avatar

Eiki Shimizu

MSc Student

Avatar

Ilana Sebag

MSc Student

Avatar

Julian Rodert

MSc Student

Avatar

Kai Biegun

MSc Student

Avatar

Kamiylah Charles

MSc Student

Avatar

Lucas Cosier

BASc Student

Avatar

Nanxi Zhang

BSc Student

Avatar

Ross Murphy

MSc Student

Avatar

Rui Li

MSc Student

Avatar

Shreshth Malik

MSc Student

Avatar

Zuka Murvanidze

BSc Student

Visitors

Affiliates

Avatar

K. S. Sesh Kumar

Research Associate

Machine learning, Discrete optimization, Differential privacy, Submodularity

Avatar

Mirgahney H. Mohamed

PhD Student

Computer vision, Uncertainty estimation

Avatar

Oscar Key

PhD Student

Probabilistic modeling, Approximate inference, Machine learning, Climate science

Alumni

Avatar

Riccardo Moriconi

PhD (10/2016-02/2021)

Machine learning, Gaussian processes, Bayesian optimization

Avatar

K. S. Sesh Kumar

Research Associate

Machine learning, Discrete optimization, Differential privacy, Submodularity

Avatar

Hugh Salimbeni

PhD (10/2015-10/2019)

Machine learning, Deep probabilistic models, Approximate inference

Avatar

Sanket Kamthe

PhD (10/2016-03/2021)

Machine learning, Reinforcement learning, Optimal control, Copulas

Avatar

Benjamin Chamberlain

PhD (10/2014-08/2018)

Machine learning, Community detection, Representation of graphs, Hyperbolic embeddings

Avatar

Simon Olofsson

PhD (06/2016-03/2020)

Machine learning, Bayesian optimization, Mechanistic models, Model discrimination

Recent Blog Posts

Matérn Gaussian Processes on Graphs

Gaussian processes are a model class for learning unknown functions from data. They are particularly of interest in statistical …

Probabilistic Active Meta Learning (PAML)

Meta-learning can make machine learning algorithms more data-efficient, using experience from prior tasks to learn related tasks …

Variational Integrator Networks

Learning models of physical systems can be tricky, but exploiting inductive biases about the nature of the system can speed up learning …

High-dimensional Bayesian Optimization Using Low-dimensional Feature Spaces

Bayesian optimization is a powerful technique for the optimization of expensive black-box functions, but typically limited to …

Aligning Time Series on Incomparable Space

Data is often gathered sequentially in the form of a time series, which consists of sequences of data points observed at successive …

Estimating Barycenters of Measures in High Dimensions

Barycenters summarize populations of measures, but computing them does not scale to high dimensions with existing methods. We propose a …

Efficiently Sampling Functions from Gaussian Process Posteriors

Efficient sampling from Gaussian process posteriors is relevant in practical applications. With Matheron’s rule we decouple the …

Healing Products of Gaussian Process Experts

Products of Gaussian process experts commonly suffer from poor performance when experts are weak. We propose aggregations and weighting …

Matérn Gaussian Processes on Riemannian Manifolds

Gaussian processes are a useful technique for modeling unknown functions. They are used in many application areas, particularly in …

Recent News

Best Student Paper Award at AISTATS 2021

Our paper on Matern Gaussian processes on graphs has been awarded the Best Student Paper Award at AISTATS

Dr. Kamthe

Dr. Sanket Kamthe successfully passed his PhD viva

Dr. Moriconi

Dr. Riccardo Moriconi successfully passed his PhD viva

Workshop on Bridging the Gap between Data-driven and Analytical Physics-based Grasping and Manipulation accepted at ICRA 2021

Workshop on Bridging the Gap between Data-driven and Analytical Physics-based Grasping and Manipulation accepted at ICRA 2021

Four Papers Acccepted at AISTATS 2021 and ICLR 2021

Four papers from our group have been accepted at AISTATS and ICLR

Recent Publications

A Practical Sparse Approximation for Real Time Recurrent Learning

Recurrent neural networks are usually trained with backpropagation through time, which requires storing a complete history of network …

Learning Contact Dynamics using Physically Structured Neural Networks

Learning physically structured representations of dynamical systems that include contact between different objects is an important …

Matérn Gaussian Processes on Graphs

Gaussian processes are a versatile framework for learning unknown functions in a manner that permits one to utilize prior information …

Aligning Time Series on Incomparable Spaces

Dynamic time warping (DTW) is a useful method for aligning, comparing and combining time series, but it requires them to live in …

Recent & Upcoming Talks

Fariba Yousefi: Multi-task Learning for Aggregated Data using Gaussian Processes

Aggregated data is commonplace in areas such as epidemiology and demography. For example, census data for a population is usually given …

Ruha Benjamin: Reimagining the Default Settings of Technology & Society

From everyday apps to complex algorithms, technology has the potential to hide, speed, and deepen discrimination, while appearing …

César Lincoln C. Mattos: Probabilistic ML-Applications and Modeling Investigations

Probabilistic modeling is a core component of modern machine learning. Nevertheless, at least from our local perspective, educational …

Jackie Kay: Fairness for Unobserved Characteristics: Insights from Technological Impacts on Que​​​​​​​er Communities

Recent advances in algorithmic fairness have largely omitted sexual orientation and gender identity. We explore the concerns of the …

Dhruva Tirumala: Using behavior priors for data efficiency in Reinforcement Learning

While reinforcement learning has shown great promise as a viable solution to many problems such as Go and Atari, it’s application …