Comparison of Time Series Models before and after Using Wavelet Shrinkage Filtering to Forecast the Amount of Natural Gas in Iraq

The procedure of reducing noise or reduction before analyzing the time series is important to get accurate and reliable outcomes when building models. Wavelet Shrinkage consisting wavelets with thresholding a powerful mathematical method used reduce that can be exposed. It has time-series observations selects cut-off threshold level suitable for removing most noise. In this paper, natural gas production data in Iraq were during (1981-2019), which included non-stationary outliers, so researcher must treat problem starting analysis. Thus, selected way analysis using Box- Jenkins (ARIMA (p,d,q) models wavelet shrinkage two methods cutting after filtering choose model defines study data. paper was concerned finding an efficient by comparing (Box-Jenkins) linear ARIMA (p, d, q) estimated from shrinkage, then decreasing rank candidate (While maintaining accuracy suitability models) re-comparing it original observations, measuring based on some statistical criteria, including AIC (Akaike information criterion), SBIC (Schwarz criterion, Bayesian criterion BIC)and HQIC (Hannan-Quinn criterion). Statistical programs such as Statgraphics XVII - X64 MATLAB analyze reached efficiency filters treating getting models, specifically filter (Daubechies(db8)) soft-threshold fixed-form method, possibility obtaining lower orders higher filtered compared corresponding is, ARIMA(0,1,1) paved Daubechies i.e., preferable do waveform concerning preference given Box (0,1,1) shrinkage. (db8)