Slight Parameter Changes Detection in Complex Signals: Multiresolution q-Entropy Automatic Tool

MARÍA EUGENIA TORRES; MARÍA MAGDALENA AÑINO; GASTÓN SCHLOTTHAUER

Baltimore, EEUU

Workshop; 2001 IEEE – EURASIP Workshop on Nonlinear Signal and Image Processing (NSIP' 2001),; 2001

IEEE- EURASIP

The continuous multiresolution entropy combines advantages stemming from both parametric q-entropy and wavelet analysis, has been shown to be sensitive to dynamical complexity changes. Classical approaches to statistical detection of such changes offer a new tool that enables their automatic detection. In this paper, a new tool for the automatic detection of slight parameter changes in nonlinear dynamical system from the analysis of the corresponding time series is proposed. The relevance of the approach, togheter with its robustness in the presence of moderate noise, is discussed in numerical simulations and it is applied to biological signals.