Using singular value decomposition to adapt Prony's method for signal denoising

Abstract

Prony’s method approximates a sequence of data points by a linear superposition of complex exponentials. The computation of the parameters of the complex exponentials by using the Moore-Penrose inverse is extended to the use of singular value decomposition (SVD) in order to obtain a dimension reduction. The application to a noise contaminated electrocardiogram signal shows the potential of the method to approximate signals by sparse spectral representations.