Signal prediction based on empirical mode decomposition and artificial neural networks

:In view of the usefulness of Empirical Mode Decomposition (EMD), Artificial Neural Networks (ANN), and Most Relevant Matching Extension (MRME) methods in dealing with nonlinear signals, we propose a new way of combining these methods to deal with signal prediction. We found the results of combining EMD with either ANN or MRME to have higher prediction precision for a time series than the result of using EMD alone.