Hefin Lambley
I am a third-year PhD student at the Mathematics Institute, supervised by Tim Sullivan.
Research interests: inverse problems • nonparametric statistics • machine learning.
Preprints
[p1] J. Bunker, M. Girolami, H. Lambley, A. M. Stuart and T. J. Sullivan (2024). Autoencoders in function space. arXiv:2408.01362
[pdf] [code]
Publications
[2] H. Lambley (2023). Strong maximum a posteriori estimation in Banach spaces with Gaussian priors. Inverse Probl. 39.12, 125010, doi:10.1088/1361-6420/ad07a4
[pdf]
[1] H. Lambley and T. J. Sullivan (2023). An order-theoretic perspective on modes and maximum a posteriori estimation in Bayesian inverse problems. SIAM/ASA J. Uncertainty Quantif. 11.4, pp. 1195–1224, doi:10.1137/22M154243X
[pdf] [supplement]
Talks and posters
- 25 Sep 2024: Autoencoders in function space. Poster at Cambridge Centre for Smart Infrastructure and Construction, Cambridge, UK (jointly presented with Justin Bunker).
- 5 Aug 2024: Autoencoders in function space. Poster at Machine Learning in Infinite Dimensions, University of Bath, UK (jointly presented with Justin Bunker).
- 13 Jun 2024: Most likely points of probability measures. Talk at the Analysis seminar, University of Augsburg, Germany.
- 5 Dec 2023: Strong maximum a posteriori estimation in Bayesian inverse problems. Talk online at A seminar on the MAP. [slides]
Education
- 2018–2022. Mathematics (MMath), University of Warwick. Class I.
- Master's dissertation: An order-theoretic perspective on modes and maximum a posteriori estimation in Bayesian inverse problems. [pdf]
Teaching
2024–2025. Teaching assistant for ES98A Fundamentals of Predictive Modelling.
2023–2024. Teaching assistant for ES98A Fundamentals of Predictive Modelling, MA264 Mathematical Methods and Modelling 3 and MA3H7 Control Theory.
2022–2023. Teaching assistant for MA133 Differential Equations.
2021–2022. Supervision of ~10 MORSE undergradute students.
Seminars
2024–2025. Organiser for the Junior Analysis and Probability Seminar (joint with Grega Saksida, Phoebe Valentine, and Pietro Wald).
