A Comparative Study between the ARIMA-Fourier Model and the Wavelet model

摘要

This work compared the performances between the autoregressive integrated moving average (ARIMA) - Fourier model and the wavelet model. The methods were applied to a consumer price index (CPI) series. The ARIMA-Fourier model was obtained by combining both the linear and sinusoidal components of the series and was found to fit adequately to the data. Similarly, the wavelet model was also diagnosed to be adequate. To fish out the most prefered model, the two competing models were subjected to comparative analysis tests. It was discovered that the wavelet model performs better than the ARIMA-Fourier model.