Recently, a Data-Driven Time-Frequency analysis method was developed to deal with nonlinear and nonstationary data. This method employs Matching Pursuit iteratively to look for the sparest representation of the signal over a huge dictionary. In this talk, I will generalize the iterative matching pursuit to some more complicated data sets, such as the data with poor scale separation, data with sparse samples and data with intra-wave frequency modulation. Moreover, the convergence of the iterative matching pursuit will also be discussed in the talk.
Copyright. All Rights Reserved.