Volume no :10, Issue no: 1, March (2019)

International Journal of Applied Mathematics and Machine Learning

ERROR ANALYSIS FOR SPARSE TIME-FREQUENCY DECOMPOSITION OF SIGNALS BASED ON PROBABILITY

Author's: Qi Yang
Pages: [47] - [57]
Received Date: March 22, 2019; Revised March 27, 2019
Submitted by: L. Rajendran.
DOI: http://dx.doi.org/10.18642/ijamml_7100122041

Abstract

In this paper, after reviewing the data driven time-frequency analysis (DDTFA) and the matching pursuit method in decomposition of intrinsic mode functions (IMFs), we estimate the orthogonality properties between IMFs based on probability. By theoretical proofs and numerical experiments, we show that the estimation in [9] is the best.

Keywords

sparse time-frequency decomposition, scale separation, nonlinear matching pursuit.