Hybrid approach Wavelet seasonal autoregressive integrated moving average model (WSARIMA) for modeling time series

Comparison of the two hybrid models, Wavelet-ARIMA and Wavelet-ES, to predict the prices of the US dollar index

“…Singh and others worked to predict the numbers of deaths from the Corona virus for the countries of Italy, America, Spain, France and Britain, by comparing the ARIMA models with the Wave-ARIMA models, and the results came to prove the preference of hybrid models, [18]. Jankova conducted an empirical study aimed at obtaining the best model for forecasting seasonal time series, by comparing SARIMA with Wave-SARIMA models, and the experimental results showed the effectiveness and accuracy of prediction for hybrid models, and that Wave-SARIMA model can be used to predict suitable seasonal time series, [6]. Finally, in a study that aims to determine the appropriate model for predicting the daily and monthly water flow in western Nishnaputna and Trinity River with different water conditions, Nourani made a comparison between non-linear models, SARIMA, ANN, ES, with hybrid models which are Wave-SARIMA, Wave-ANN, Wave -ES, as well as a proposed hybrid model which is Wave-WES, the results of the comparison indicated that the hybrid models are better in reducing the impact of noise and seasonal changes in the time series than the traditional SARIMA models, and that the proposed hybrid model Wave-WES is the most suitable model for predicting water flow , [11].…”