Date of Submission

Spring 2012

Academic Program

Mathematics

Project Advisor 1

Cliona Golden

Abstract/Artist's Statement

The Empirical Mode Decomposition (EMD) is a signal processing technique designed to express nonstationary and nonlinear data as a sum of a small number of components called intrinsic mode functions (IMFs) and a monotonic residual term. These IMFs contain oscillations in a narrow band of scale, have symmetric envelopes, are roughly zero mean and have other regularity properties. A new approach to filling large gaps in oscillatory datasets by applying EMD and gap-filling the IMFs is examined. A novel means of setting the parameters of the method adaptively is developed and explicated. Quantitative and statistical means are brought to bear to demonstrate by simulation that the method is effective. It is shown that for data exhibiting deep structure and long-range dependence, EMD gap-filling is superior to linear interpolation and succeeds in catching the oscillations present in the missing data.

Distribution Options

Access restricted to On-Campus only

Creative Commons License

Creative Commons License
This work is licensed under a Creative Commons Attribution 3.0 License.

Share

COinS