WSIC Criterion for Decomposition Level Selection of Orthogonal Wavelet Transform

Wavelet transform is a widely used method in the field of signal analysis, and its application effect depends largely on the selection of wavelet decomposition level. In view of the unknown original signal, this paper transforms the selection of wavelet decomposition level into a model selection problem through statistical modeling, and then analyzes and presents wavelet selection information criteria (WSIC) of orthogonal wavelet transform level selection from the perspective of model selection. Finally, simulation experiments are conducted to verify and compare the effect of the WSIC information criteria with AIC and BIC criteria on wavelet level selection. The results show that the level accuracy of WSIC is up to 15.1% higher than that of AIC criterion and up to 14.3% higher than that of BIC criterion, indicating that WSIC criterion has better stability in selecting wavelet decomposition level than that of AIC and BIC criterion in existing literatures.