Statistical Machine Learning Group

University College London

The Statistical Machine Learning group is 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.

If you are interested in joining the Statistical Machine Learning group, please check out our openings.

Recent & Upcoming Talks

Michael Lutter: Inductive Biases for Learning Robot Control

In order to leave the factory floors and research labs, future robots must abandon their stiff and pre-programmed movements and be …

2020-02-20 15:00 — 16:00 AI Centre, UCL

Marc Deisenroth

Vincent Adam: Doubly Sparse Variational Gaussian Processes

The use of Gaussian process models is typically limited to datasets with a few tens of thousands of observations due to their …

2019-10-31 10:00 — 11:00 AI Centre, UCL

Marc Deisenroth

PDF

Samuel Kaski: Probabilistic Modelling with Experts

I will discuss multiple-data-source prediction and modelling problems arising in a number of fields, for instance in omics-based …

2019-10-04 10:30 — 11:30 LT 139 (Huxley), Imperial College London

Marc Deisenroth

PDF

Christian Walder: New Tricks for Estimating Gradients of Expectations

We derive a family of Monte Carlo estimators for gradients of expectations, which is related to the log-derivative trick, but involves …

2019-09-30 11:00 — 12:00 LT 130 (Huxley), Imperial College London

Marc Deisenroth

PDF

Emtiyaz Khan: Learning-Algorithms from Bayesian Principle

In machine learning, new learning algorithms are designed by borrowing ideas from optimization and statistics followed by an extensive …

2019-09-16 14:00 — 15:00 Huxley LT 144, Imperial College London

Riccardo Moriconi, Marc Deisenroth

Slides

People

Principal Investigators

Marc Deisenroth

DeepMind Chair in Artificial Intelligence

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

Postdoctoral Researchers

K. S. Sesh Kumar

Research Associate

Machine learning, Discrete optimization, Differential privacy, Submodularity

PhD Students

Alexander Terenin

PhD Student

Machine learning, Bayesian statistics, Geometric learning, Meta learning

James Wilson

PhD Student

Machine learning, Bayesian optimization, Practical approximate inference

Janith Petangoda

PhD Student

Machine learning, Meta learning, Differential geometry, Reinforcement learning

Riccardo Moriconi

PhD Student

Machine learning, Bayesian optimization

Samuel Cohen

PhD Student

Machine learning, Optimal transport

Sanket Kamthe

PhD Student

Machine learning, Reinforcement learning, Optimal control, Copulas

Simon Olofsson

PhD Student

Machine learning, Bayesian optimization, Mechanistic models, Model discrimination

Steindór Sæmundsson

PhD Student

Machine learning, Meta learning, Structural priors

Visitors

Rasmus Larsen

Intern

Alumni

Benjamin Chamberlain

PhD (10/2014 - 08/2018)

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

Hugh Salimbeni

PhD (10/2015-10/2019)

Machine learning, Deep probabilistic models, Approximate inference

Featured Publications

Steindór Sæmundsson, Alexander Terenin, Katja Hofmann, Marc P. Deisenroth

2020-06-04 Proceedings of the International Conference on Artificial Intelligence and Statistics (AISTATS)

Variational Integrator Networks for Physically Meaningful Embeddings

Learning workable representations of dynamical systems is becoming an increasingly important problem in a number of application areas. By leveraging recent work connecting deep neural networks to systems of differential equations, we propose variational integrator networks, a class of neural network architectures designed to ensure faithful representations of the dynamics under study. This class of network architectures facilitates accurate long-term prediction, interpretability, and data-efficient learning, while still remaining highly flexible and capable of modeling complex behavior. We demonstrate that they can accurately learn dynamical systems from both noisy observations in phase space and from image pixels within which the unknown dynamics are embedded.

PDF

Marc P. Deisenroth, A. Aldo Faisal, Cheng Soon Ong

2020-03-01 Cambridge University Press

Mathematics for Machine Learning

Mathematics for Machine Learning is a book that motivates people to learn mathematical concepts. The book is not intended to cover advanced machine learning techniques, because there are already plenty of books doing this. Instead, we aim to provide the necessary mathematical skills to read those other books.

PDF

Gianfranco Bertone, Marc P. Deisenroth, Jong S. Kim, Sebastian Liem, Roberto Ruiz de Austri, Max Welling

2019-03-01 Physics of the Dark Universe

Accelerating the BSM Interpretation of LHC Data with Machine Learning

The interpretation of Large Hadron Collider (LHC) data in the framework of Beyond the Standard Model (BSM) theories is hampered by the need to run computationally expensive event generators and detector simulators. Performing statistically convergent scans of high-dimensional BSM theories is consequently challenging, and in practice unfeasible for very high-dimensional BSM theories. We present here a new machine learning method that accelerates the interpretation of LHC data, by learning the relationship between BSM theory parameters and data. As a proof-of-concept, we demonstrate that this technique accurately predicts natural SUSY signal events in two signal regions at the High Luminosity LHC, up to four orders of magnitude faster than standard techniques. The new approach makes it possible to rapidly and accurately reconstruct the theory parameters of complex BSM theories, should an excess in the data be discovered at the LHC.

PDF DOI

James T. Wilson, Frank Hutter, Marc P. Deisenroth

2018-12-01 Advances in Neural Information Processing Systems (NeurIPS)

Maximizing Acquisition Functions for Bayesian Optimization

Bayesian optimization is a sample-efficient approach to global optimization that relies on theoretically motivated value heuristics (acquisition functions) to guide its search process. Fully maximizing acquisition functions produces the Bayes’ decision rule, but this ideal is difficult to achieve since these functions are frequently non-trivial to optimize. This statement is especially true when evaluating queries in parallel, where acquisition functions are routinely non-convex, high-dimensional, and intractable. We first show that acquisition functions estimated via Monte Carlo integration are consistently amenable to gradient-based optimization. Subsequently, we identify a common family of acquisition functions, including EI and UCB, whose characteristics not only facilitate but justify use of greedy approaches for their maximization.

PDF Code

Steindór Sæmundsson, Katja Hofmann, Marc P. Deisenroth

2018-08-01 Proceedings of the Conference on Uncertainty in Artificial Intelligence (UAI)