Comparison of Some Wavelet Transformations to Estimate Nonparametric Regression Function

The purpose of this article is to improve and minimize noise from the signal by studying wavelet transforms and showing how to use the most effective ones for processing and analysis. As both the Discrete Wavelet Transformation method was used, we will outline some transformation techniques along with the methodology for applying them to remove noise from the signal. Proceeds based on the threshold value and the threshold functions Lifting Transformation, Wavelet Transformation, and Packet Discrete Wavelet Transformation. Using AMSE, A comparison was made between them , and the best was selected. When the aforementioned techniques were applied to actual data that was represented by each of the prices, it became evident that the lifting transformation method (LIFTINGW) and the discrete transformation method with a soft threshold function and the Sure threshold value (SURESDW) were the best. Consumer prices will be the dependent variable for the period of 2015–2020, and Iraqi oil (Average price of a barrel of Iraqi oil) will serve as the explanatory variable. The methods described above have proven effective in estimating the nonparametric regression function for the study model. Paper type: Research paper.