The Combination of ARIMA with Wavelet De-Noise

The ARIMA model has the ability to extract the useful data from the originate data and the remain is the noise data. But if the originate data has a large amount of noise data, this will has great influence in ARIMA model such as the parameters. In order to eliminate the effect of the noise, this paper combine the ARIMA model with wavelet denoise. First, the originate data is denoise through wavelet transform. Here a reasonable threshold is given according to special method. Then the ARIMA model can be used to work on data analysis. The application result shows that this method can work perfectly.