Data-driven signal decomposition method

This paper introduces the data-driven signal decomposition method based on the Empirical Mode Decomposition (EMD) technique. The decomposition process uses the data themselves to derive the base function in order to decompose the one-dimensional signal into a finite set of intrinsic mode signals. The novelty of EMD is that the decomposition does not use any artificial data windowing which implies fewer artifacts in the decomposed signals. The results show that the method can be effectively used in analyzing non-stationary signals. Furthermore, we applied this method to analyze closing equity prices of a financial stock. The result demonstrates the usefulness of the method in analyzing financial time series data, and some practical considerations in envelope estimation.