University of Göttingen - Institute for Numerical and Applied Mathematics
Research group for Mathematical Signal and Image Processing


Prony like methods for noisy data



(09/2015 - 02/2017)

Solitoneneigenschaften
Individual research grant of Chinese Scholarship Council

We investigate the stability of Prony's method in case of noisy input data. For this purpose we consider a maximum likelihood modification of the Prony method and compare it to other approaches for stabilization of Prony's method. We further aim at generalizing the approach to sparse expansions of eigenfunctions of linear operators.



Principal investigator: Gerlind Plonka-Hoch, (supervisor)
Staff: Ran Zhang

Corresponding publications
Optimal approximation with exponential sums by maximum likelihood modification of Prony's method
Ran Zhang, Gerlind Plonka
Advances in Computational Mathematics 45(3), 1657-1687, 2019, revised preprint as download.



Research Group for Mathematical Signal and Image Processing

Institute for Numerical and Applied Mathematics
Lotzestr. 16-18
37083 Göttingen