Jeremy Levesley's Homepage
Currently I am interested in approximation in Euclidean space and on spheres using radial basis functions, and generalisations of these procedures to locally compact manifolds. I am interested in the applications of RBFs in finance, especially practical high dimensional approximation using sparse grid methods. I am also involved in forensic applications with fingerprints. I am also interested in Smart City and Digital Medicine. Current projects include
- Sparse kernel approximation with Manolis Georgoulis, Xingping Sun, Alex Kushpel and PhD students Peter Dong and Fuat Usta;
- lattice Boltzmann methods for modelling compressible fluid flow (EPSRC funded RA Robert Brownlee); code and pictures available here;
- Modelling of methane hydrate dissociation (NERC funded post doc Denis Goldobin, Nikolai Brilliantov, Mike Lovell (Leicester), John Rees (BGS)).
- Radial basis function applications in pricing derivatives with PhD students Juxi Li and Yangzhang Zhao.
- Developing surrogate models in engineering, in partnership with Alstom, and PhD student Qi Zhang.
- Image processing of fingerprints as part of the INTREPID FP7 project, with PhD student Etienne Pillin.
- Approximation on manifolds with Alexander Kushpel.
National Teaching Fellow 2014
Having done a P.G.C.E. I retain a big interest in enabling the learning of our students, both undergraduate and postgraduate. My main efforts have been directed towards increasing the students desire to own what they learn. To this end I have developed the Peer Support (PS) system (based on Supplemental Instruction, a scheme originating with UMKC), in which more experienced students facilitate groups of first year students. The main aim is for the first year students to pool their knowledge in order to gain a richer understanding of a particular idea. The more experienced student uses their experience to give the first years strategies by which they may understand a particular idea.