The Best Mother Wavelet for Stock Prediction using Adaplet Method

摘要

It has been determined the best mother wavelet for stock prediction of SONY in 2006 and BNI in 2012 based on the Adaplet Method (the adaptive filter which uses wavelet as initial coefficients). The Mother Wavelets are Coiflet 1-5, Daubechies 1-5, and Symlet 1-5. The prediction analysis includes overshoot, autocorrelation error and data pattern conformity, three days prediction, and segmentation. According to the overshoot analysis, it shown that for all data, the overshoot at the beginning of data increased as its wavelet level increased, while Daubechies 1 and Symlet 1 produced smallest overshoot among other wavelets (112.2%). The autocorrelation error of data pattern prediction indicates conformity with the original data. As its wavelet level increased, the autocorrelation error pattern also ramped (near zero). The Coiflet 5 and the Daubechies 1 resulted in smallest mean square error (MSE) of about 0.0147, while the Coiflet 1 resulted in the best result with an average error of 0.001 for the analysis of next three days prediction. On the other hands, Symlet 3 resulted in the best MSE of 1.213. Overall, Symlet offers the best result, according to best wavelet sequence assessment of each method.