De-noising with wavelets method in chaotic time series: application in climatology, energy, and finance (Invited Paper)
Dominique Guégan, kebira Hoummiya · Proceedings of SPIE, the International Society for Optical Engineering/Proceedings of SPIE · 2005
In this paper, in order to de-noise a chaotic signal, we compare the time-frequency deconvolution method with the wavelets method. We apply our results on different dynamical systems and show the capability of wavelets' method to reconstruct the attractor of a chaotic time series. Then, we de-noise different data sets in order to re-built their attractor using the wavelets method. The applications concern temperatures and wind fluctuations, electricity spot prices and financial data sets.