In this study, the enhanced hybrid empirical mode decomposition with autoregressive integrated moving average (EMD- ARIMA) method is proposed and applied to forecast daily new COVID-19 reported cases in Jordan. The EMD method is applied to decompose the COVID-19 data into a number of IMFs components as a simple time series. Then, the appropriate ARIMA(p,d,q) model is applied to evaluate the forecasting value for the low-frequency components. Then, the forecasting results are collected together. Data for this study are collected from the Jordanian Ministry of Health. Seven forecasting accuracy measures are employed to compare the forecasting results of the proposed technique with the results of seven forecasting methods. The comparison of forecasting results shows that the enhanced EMD-ARIMA method is better than selecting forecasting methodologies in COVID-19 data.
Digital Object Identifier (DOI)
M. Awajan, Ahmad; Al-Hasanat, Bilal; Elkaroui, Elarbi; AL e’damat, Ayed; S. Al-Gounmeein, Remal; and S. Al- Jawarneh, Abdullah
"Time Series Forecasting of New Cases for COVID-19 Pandemic in Jordan Using Enhanced Hybrid EMD-ARIMA,"
Journal of Statistics Applications & Probability: Vol. 13:
1, Article 18.
Available at: https://digitalcommons.aaru.edu.jo/jsap/vol13/iss1/18