The Wavelet Transformation and Decomposition of ARIMA Models

Abstract

Abstract. The objective of this article is to study the effect of wavelet filter on the time series data. By usingwavelet transformation and hard thresholding technique, the ARIMA model decomposed into sum of two ARIMAmodels and the relation between sum of square errors due to the ARIMA model and sum of two ARIMA modelswill be discussed. . By using wavelet transformation and soft thresholding technique, the ARIMA modeldecomposed into sum of three ARIMA models and the relation between sum of square errors due to the ARIMAmodel and sum of three ARIMA models will be discussed.