EMD and EEMD has been used on time series across diverse fields (Biomedical, Geology, Seismology, Mechanical vibrations, Financials). Most of the articles have more information on applying EMD but not clearly articulating the real purpose and how they are using the information derived from the IMFs for their end goals.

Following are the key gaps that I am looking for.

1. Why EMD/EEMD is required for time series?

2. What useful information can be derived from IMFs that cannot be extracted in the original series

3. When to go for EMD/EEMD for any time series?

4. Where can EMD/EEMD be useful - nonstationary, nonlinear, nondeterministic conditions

5. After computing IMFs (First 3, 5 or N), how to select the most useful IMF? And what is the use of comparing features of IMF to the features extracted from original series.

Any good paper on EMD/EEMD or reference book on this topic will be helpful.

Thanks in advance