Team

Mark Girolami is the Sir Kirby Laing Professor of Civil Engineering (1965) within the Department of Engineering at the University of Cambridge where he also holds the Royal Academy of Engineering Research Chair in Data Centric Engineering. In this role he will provide academic leadership for the Centre for Digital Built Britain across the University and more broadly throughout the national and international research communities. Prior to joining the University of Cambridge he held the Chair of Statistics in the Department of Mathematicsat Imperial College London.

He was one of the original founding Executive Directors of the Alan Turing Institute the UK’s national institute for Data Science and Artificial Intelligence, after which he was appointed as Strategic Programme Director at Turing, where he established and continues to lead the Lloyd’s Register FoundationProgramme on Data Centric Engineering.

Professor Girolami is an elected fellow of the Royal Society of Edinburgh, he was an EPSRC Advanced Research Fellow (2007-2012), an EPSRC Established Career Research Fellow (2012-2018), and a recipient of a Royal Society Wolfson Research Merit Award.

He delivered the IMS Medallion Lecture at the Joint Statistical Meeting 2017, and the Bernoulli Society Forum Lecture at the European Meeting of Statisticians 2017.

In 2020 Professor Girolami will deliver the BCS and IET Turing Talk in London, Manchester, and Belfast.

Professor Girolami currently serves as the Editor-in-Chief of Statistics and Computing and the new open access journal Data Centric Engineeringpublished by Cambridge University Press.

Mike Giles is a Professor of Scientific Computing at the University of Oxford. His research focuses on improving the accuracy, efficiency and analysis of Monte Carlo methods. A particular highlight has been the development and numerical analysis of multilevel Monte Carlo methods which are now being used widely. He is interested in high performance parallel computing, and for almost 10 years has been working on the exploitation of GPUs (graphics processors) for a variety of financial, scientific and engineering applications. His research in ICONIC will cover both of these aspects, looking at the development and analysis of efficient Monte Carlo solution methods for the models developed by Mark and Des, and also mixed precision and parallel computing implementations on GPUs and other many-core processors such as Intel’s latest Xeon CPUs.

Des Higham is the 1966 Chair of Numerical Analysis at the University of Edinburgh. He is an EPSRC/RCUK Digital Economy Established Career Fellow. Des is a Fellow of the Royal Society of Edinburgh, a SIAM Fellow, and recently held a Royal Society Wolfson Research Merit Award. Des has experience in stochastic modelling and computation, and has been involved in extending the scope of Mike’s multilevel Monte Carlo techniques to discrete space models. He will be working with Mark on modelling and inference for urban behaviour, notably crime, and with Nick on some of the underlying issues in matrix computation.

Nick Higham is Royal Society Research Professor and Richardson Professor of Applied Mathematics at the University of Manchester. His research focuses on the development of algorithms, primarily in numerical linear algebra, and analysis of their accuracy and stability. In ICONIC his research topics will include functions of matrices, rounding error analysis, and matrix nearness and completion problems. In particular, he will work with Mike to develop effective mixed precision algorithms that exploit GPUs other many-core processors. Nick is a Fellow of the Royal Society, a SIAM Fellow, and a Member of Academia Europaea. He publishes a blog about applied mathematics.

Research Associates

Matteo Croci (University of Oxford 2020-) works within the fields of computational stochastics, uncertainty quantification and industrial mathematics, with a focus on multilevel (quasi-) Monte Carlo methods, high performance computing, partial differential equations with random coefficients and the finite element method. Matteo’s research interests also include brain modelling, biomedical computing and deflation techniques for variational inequalities. During his DPhil (PhD) at the University of Oxford, Matteo developed two new state-of-the-art multilevel (quasi-) Monte Carlo techniques, which he employed to quantify the uncertainty in numerical simulations of brain fluid and solute movement in real-life geometries. More recently within ICONIC he has analysed the effect of rounding errors on the approximation of parabolic PDEs.

Ieva Kazlauskaite (University of Cambridge) is a research associate working on stochastic modelling of urban structure. Her research interests to date include the application of Gaussian processes and other Bayesian nonparametric techniques to the study of time-series data.

Henry-Louis de Kergorlay (University of Edinburgh 2020-) obtained his PhD from the University of Edinburgh; during this time he spent a year at the Alan Turing Institute as an Enrichment Student. He has research interests in analysis, random graphs and topology, and their applications in data science. His current activities include modeling, analysis and simulation of epidemics in a hypergraph setting; this accounts for interactions at the group level rather than the traditional pairwise case.

Mantas Mikaitis (University of Manchester 2020-) obtained a BSc(Hons) and a PhD in computer science from the University of Manchester in 2016 and 2020 respectively. In his studies he worked on the SpiNNaker project led by Prof. Steve Furber and, under the supervision of David R. Lester, explored numerical accuracy of neural simulations on the SpiNNaker neuromorphic chip and has developed arithmetic accelerators for the next generation SpiNNaker2 chip. In 2019 he received a prestigious EPSRC Doctoral Prize Fellowship as part of which he is researching low precision arithmetic in numerical linear algebra. His research interests include various aspects of computer arithmetic and related topics.

Srikara Pranesh (University of Manchester 2019-) obtained his PhD from the department of Civil Engineering at the Indian Institute of Science, Bangalore in 2018. During his PhD his research focused on developing efficient numerical algorithms for large scale uncertainty quantification. Currently his research interests include High Performance Computing, Multi-Precision Algorithms, and numerical aspects of Uncertainty Quantification.

Kangrui Wang (University of Cambridge 2020-) obtained his PhD from the department of Mathematics at the University of Leicester in 2018. His research focus on stochastic modelling of high-dimensional data and developing non-stationary non-separable kernels for Gaussian process based models. His interest also lies in the application of deep Gaussian processes to non-stationary spatio-temporal problems.  Before joining this project, he was a research associate in the Alan Turing Institute under London air quality project. 

Associated Research Students

Michael Connolly (University of Manchester 2018-) is a third year PhD student working with Professor Nick Higham. His research focuses on developing a new generation of numerical linear algebra algorithms that exploit current and future computers. Prior to joining the group, Michael obtained a degree in Theoretical Physics from Trinity College Dublin. He also has a keen interest in Machine Learning and Deep Learning.

Jan Povala (Imperial College London) is a DPhil student working on crime prediction and statistical inference using spatio-temporal log-Gaussian Cox processes.  His project is part of the Financial Computing CDT at Imperial College London.

Oliver Sheridan-Methven (University of Oxford) is a DPhil student working on the use of high performance low/mixed precision vectorised computer arithmetic for stochastic simulations.  His project is sponsored by ARM and NAG and is part of the InFoMM CDT.

Former Research Associates

Pierre Blanchard (University of Manchester 2017-2019) obtained his PhD in Applied Mathematics and Scientific Computing from the University of Bordeaux, France, in 2017. A theme of his research to date has been hierachical representation of matrices, as applied to various domains of scientific computing, including seismic wave scattering, materials sciences and geostatistics. He has left the ICONIC project and is now working at Arm.

Louis Ellam (Imperial College London) was a research assistant working on stochastic modelling and statistical inference for urban systems and is completing his PhD in Statistics at Imperial College London.  His research to date has focused on the development of new Bayesian methodologies for inverse problems, and has included high-dimensional Gaussian models, elliptic PDEs and singly-constrained dynamic urban models. He had now taken up a lectureship at DeMontfort University and left the ICONIC project.

Wei Fang (University of Oxford 2018-2019) was a research assistant who completed his DPhil (PhD) in Mathematics at the University of Oxford. His research focused on the development of a new Multilevel Monte Carlo method using an adaptive approximation of the solution to stochastic differential equations with non-globally Lipschitz drift over a finite time interval, as well as computing expectations and sensitivities with respect to the invariant measure for ergodic SDEs.

Craig Gilmour (University of Strathclyde 2018-2020) is a research assistant working on crime prediction, and is completing his PhD in Mathematics at University of Strathclyde. His research to date has focused on applying self-exciting point processes in the context of urban crime.

Theo Mary (University of Manchester 2018-2019) obtained his PhD in Applied Mathematics and Computer Science from the University of Toulouse, France, in 2017. His research to date has focused on exploiting low-rank approximations inside direct, factorization-based methods to reduce the computational cost of the solution of sparse linear systems arising in several applications, including seismic imaging, electromagnetic modeling, and structural mechanics. He is now CNRS researcher at LIP6 laboratory in Paris, France.

Seppo Virtanen (The University of Cambridge 2018-2020) works at the intersection between statistical inference and mathematical modeling. His research interests include multi-view learning, latent variable modeling and approximate Bayesian inference.  His recent work focuses on dynamic models of text documents. He is now a Lecturer in Statistics in the Department of Mathematics at the University of Leeds