Abdullah Almasri: Forecasting Long Memory Processes Using Wavelet Transform

This study outlines a framework for forecasting long memory processes using discrete wavelet transform (DWT) based multiresoulution analysis (MRA). We compare forecasted values obtained from modeling the data in the wavelet domain with the forecasted values from the time domain. The results from the simulation study have shown that the DWT forecasting is more accurate than the forecasting under the time domain.