Benjamin M. Kent

I am currently a Research Associate in the Machine Learning Group in the Department of Engineering, University of Cambridge.

Previously, I was a Assegnista di Ricerca at Istituto di Matematica Applicata e Tecnologie Informatiche “E. Magenes” (IMATI) working with Lorenzo Tamellini on multi-fidelity surrogate modelling, with a focus on uncertainty quantification.

I attained a PhD in Numerical Analysis from the Department of Mathematics at The University of Manchester. I was supervised by Professor Catherine Powell and Professor David Silvester. It was an iCASE sponsored project with IBM Research UK.

I attained a First-Class Honours BSc in Mathematics and Physics BSc at the University of Warwick.

Between my undergrature and PhD degrees, I spent four years in industrial research at the multi-national engineering company Thales UK.

Contact

Email: ben kent AT live DOT co DOT uk
Website: https://benmkent.github.io/
GitHub: https://github.com/benmkent/
LinkedIn: https://www.linkedin.com/in/benjaminmkent/
ORCiD: https://orcid.org/0000-0003-4968-7993

Experience

Research Associate
Department of Engineering, University of Cambridge, Cambridge, UK
September 2025 – Present

Investigating probabilistic machine learning techniques as part of the Prob_AI Hub.

Assegnista di Ricerca (Postdoctoral Researcher)
Istituto di Matematica Applicata e Tecnologie Informatiche “E. Magenes”, Pavia, Italy
February 2024 – September 2025

Developed novel multi-fidelity surrogate modelling algorithms for use with “noisy” solution data, including applications to complex, coupled modelling problems.
Focused on data-driven approximation via sparse grid polynomial interpolation methods and exploiting spectral polynomial approximation properties. We used adaptive experimental design optimally select sampling points to fit a multi-fidelity surrogate model. Our novel algorithm offers robust computationally efficient surrogate modelling for parametric problems. We focused on applications in uncertainty quantification and optimisation.
Implemented and analysed complex test problems including a NASA turbulent fluid flow validation case via OpenFOAM and ParaView, and use of SLURM high performance computing facilities.
Numerical experiments required containerisation and deployment of Python, Julia, and more complex solvers via Docker and Kubernetes. This included contributions to benchmark models for the Democratising Uncertainty Quantification project.
Contributed to the development of the widely used Sparse Grids MATLAB Kit and developed a novel Julia surrogate modelling package SparseGridsKit.jl.
Co-organised mini-symposia focusing on surrogate modelling at Adaptive Modeling and Simulation (ADMOS) 2025 and the Biennial Numerical Analysis Conference 2025.
Presented a seminar at Sandia National Laboratories, USA, and an invited talk for the UM-Bridge Workshop 2024 in addition to a number of conference talks.
Supervised by Lorenzo Tamellini, with collaborators at Universitat Politècnica de Catalunya, Spain.
IMATI is a Consiglio Nazionale delle Ricerche (CNR) institute.

PhD in Numerical Analysis
The University of Manchester, UK
September 2019 – November 2023

Industry sponsored ICASE project with IBM Research UK, supervised by Professor Catherine Powell and Professor David Silvester.
Investigated computationally efficient adaptive-in-time approximation for parametric, time-dependent partial differential equations.
Specifically, developed a novel adaptive-in-time sparse-grid stochastic collocation algorithm for approximation of parametric time-dependent advection–diffusion problems. Numerical experiments demonstrated the computational efficiency even for high dimensional parametric inputs that would be intractable for naive tensor-based approximations.
Studied topics in Functional Analysis, Approximation Theory and Finite Element Analysis, Adaptive Finite Element Methods, Uncertainty Quantification (Monte Carlo methods, stochastic collocation, stochastic Galerkin methods) and Bayesian inverse problems.
Co-organiser of Manchester Mathematics Research Student Conference in 2020 and Mathematics of Data Science student conference in 2020.
Communication of my research included an invited seminar for IBM Research UK in 2022 and “Best Student Talk” prize at the Manchester SIAM-IMA Student Chapter Conference 2023.
Thesis: Efficient Approximation of Parametric Parabolic Partial Differential Equations.
Teaching assistant for Matrix Analysis MATH36001, Mathematical Workshop MATH10001, and Complex Analysis MATH20142.

Algorithm Developer
Thales, Stockport, UK
September 2017 – August 2019

Developed array signal processing algorithms in MATLAB for time-series sensor data. Tested and evaluated algorithms on large data sets from customer experiments.
Collaborated with systems engineers to transform customer requirements to novel algorithm specifications. In particular, worked on building realistic simulator tools for user training and software testing.
Worked alongside software engineers to implement and validate novel algorithms in delivered product.
Post event analysis, data cleaning, and reporting for newly collected experimental data sets.

Research Engineer
Thales, Reading, UK
September 2015 – September 2017

Completed a two-year graduate scheme with training in both technical and soft skills.
Four project placements:
- implementation of state of the art cryptographic key exchange algorithms in C for evaluation,
- fast prototyping of radar signal processing algorithms in MATLAB and SIMULINK,
- data filtering, multi-target data fusion and tracking algorithms in MATLAB,
- investigation of environmental effects on array signal processing algorithms and experimental data analysis.

Mathematics and Physics BSc, First-Class Honours
The University of Warwick, UK
September 2012 – July 2015

Prize for the best exam results in my cohort.
Studied modules include Matrix Analysis & Algorithms, Numerical Analysis & PDEs, Scientific Programming, C Programming.

Rules and Procedures Software Internship
Lloyd’s Register, Southampton, UK
July 2014 – August 2014

Upgraded finite element approximation software components from FORTRAN to C++, in particular algorithms for element shape validation.

The College of Richard Collyer, Horsham, UK
September 2010 – July 2012

A Levels: Mathematics A*, Further Mathematics A*, Physics A*, Chemistry A*, Electronics A*.
Previously, GCSE: 10 A* (including Maths and English) + 1 A (French).

Publications

B. M. Kent, L. Tamellini, M. Giacomini, and A. Huerta, Noise-robust multi-fidelity surrogate modelling for parametric partial differential equations, arXiv preprint, arXiv:2507.03691 (2025), https://doi.org/10.48550/arXiv.2507.03691.
B. M. Kent, SparseGridsKit.jl: Adaptive single- and multi-fidelity sparse grid approximation in Julia, J. Open Source Software, 10 (113) (2025), 8300, https://doi.org/10.21105/joss.08300.
L. Seelinger, A. Reinarz, M. B. Lykkegaard, R. Akers, A. M. A. Alghamdi, D. Aristoff, W. Bangerth, J. Bénézech, M. Diez, K. Frey, J. D. Jakeman, J. S. Jørgensen, K.-T. Kim, B. M. Kent, M. Martinelli, M. Parno, R. Pellegrini, N. Petra, N. A. B. Riis, K. Rosenfeld, A. Serani, L. Tamellini, U. Villa, T. J. Dodwell, and R. Scheichl, Democratizing uncertainty quantification, J. Comput. Phys., 113542 (2024), https://doi.org/10.1016/j.jcp.2024.113542.
B. M. Kent, C. E. Powell, D. J. Silvester, and M. J. Zimoń, Efficient adaptive stochastic collocation strategies for advection–diffusion problems with uncertain inputs, J. Sci. Comput., 96 (2023), 64, https://doi.org/10.1007/s10915-023-02247-w.
B. M. Kent, Efficient approximation of parametric parabolic partial differential equations, Ph.D. thesis, University of Manchester, 2024.

Talks

30th Biennial Numerical Analysis Conference, June 2025. Noise Robust Multi-Index Stochastic Collocation for Parametric PDEs.
XII International Conference on Adaptive Modeling and Simulation (ADMOS), June 2025. Noise Robust Multi-Index Stochastic Collocation for Parametric PDEs.
Sandia National Laboratories, March 2025. Multi-Index Stochastic Collocation for PDEs with Imperfect Solvers.
SIAM Conference on Computational Science and Engineering, March 2025. Multi-Index Stochastic Collocation for PDEs with Imperfect Solvers.
UM-Bridge Workshop, December 2024. An UM-Bridge-based setup for multi-fidelity surrogate models for UQ (invited talk). [PDF] [YouTube]
CNR-IMATI Internal Conference, November 2024. Multi-fidelity Approach for Uncertainty Quantification of a Fluid Dynamics NASA Test Case.
Workshop on Frontiers of Uncertainty Quantification, September 2024. Adaptive Stochastic Collocation for Parametric Parabolic PDEs.
Very Informal Seminar Series, University of Pavia, May 2024. Computationally Efficient Approximation of Parametric Partial Differential Equations.
29th Biennial Numerical Analysis Conference, June 2023. Adaptive in Time Approximation of Parametric Parabolic PDEs.
Manchester SIAM-IMA Student Chapter Conference, April 2023. Adaptive in Time Approximation of Parametric Parabolic PDEs (Best Student Talk Prize Winner). [PDF]
SIAM Conference on Computational Science and Engineering, February 2023. Efficient Adaptive Stochastic Collocation Strategies for Advection-Diffusion Problems with Uncertain Inputs.
SIAM UKIE National Student Chapter Conference, June 2022. Error Estimation for Stochastic Collocation Approximation of Parametric Advection–Diffusion Problems.
IBM Research UK, April 2022. Efficient Approximation of Parametric Parabolic PDEs (invited seminar).
SIAM Conference on Uncertainty Quantification, April 2022. A Posteriori Error Estimation for Stochastic Collocation Applied to Parametric Parabolic PDEs.
26th Annual Meeting of SIAM UKIE Section, January 2022. A Posteriori Error Estimation for Stochastic Collocation Applied to Parametric Parabolic PDEs.

Organisation

Co-organiser: Minisymposium on Recent Advances in Multifidelity Uncertainty Quantification, SIAM Conference on Uncertainty Quatification, March 2026.
Co-organiser: Minisymposium on Approximating complex systems: Surrogates, reduced order modelling and dimension reduction *, 30th Biennial Numerical Analysis Conference, June 2025.
Co-organiser: Minisymposium on Adaptive sampling and surrogate/reduced order modelling strategies for parametric differential equations, XII International Conference on Adaptive Modeling and Simulation (ADMOS), June 2025.
Co-organiser: Manchester Mathematics Research Student Conference online conference, 2020.
Co-organiser: Mathematics of Data Science online student conference, 2020.

Contact

Experience

Research AssociateDepartment of Engineering, University of Cambridge, Cambridge, UKSeptember 2025 – Present

Assegnista di Ricerca (Postdoctoral Researcher)Istituto di Matematica Applicata e Tecnologie Informatiche “E. Magenes”, Pavia, ItalyFebruary 2024 – September 2025

PhD in Numerical AnalysisThe University of Manchester, UKSeptember 2019 – November 2023

Algorithm DeveloperThales, Stockport, UKSeptember 2017 – August 2019

Research EngineerThales, Reading, UKSeptember 2015 – September 2017

Mathematics and Physics BSc, First-Class HonoursThe University of Warwick, UKSeptember 2012 – July 2015

Rules and Procedures Software InternshipLloyd’s Register, Southampton, UKJuly 2014 – August 2014

The College of Richard Collyer, Horsham, UKSeptember 2010 – July 2012