Evolving best-basis representations
A.R. Ferreira da Silva · 2003
Adapted waveform analysis decomposes a signal or dataset into a linear expansion of waveforms that are selected from a redundant dictionary of functions. These waveforms are chosen in order to best match the signal structures. We consider decompositions using libraries of time shifted wavelet packet bases. We use evolutionary programming as the main tool to solve the optimization problem of finding the joint best basis, best shift and best wavelet filter for representing a given signal. The evolved best basis is combined with linear discriminant analysis for the purpose of temporal pattern identification.