Algorithm 1: Pseudo-code for wavelet-based feature creation

Algorithm To Construct Features Vector

Input: recorded SKT, EBT, and TSVs Output: table with two more features for each signal, and TSVs Begin:    read data table;    FOR each SKT:        IF (SKT is outlier OR null):        remove the row;        END IF;    END FOR;    FOR each EBT:        IF (EBT is outlier OR null):        remove the row;        END IF;    END FOR;    denoise SKT and EBT using MATLAB function;    decompose SKT and EBT using wavelet decomposition function;    reconstruct SKT approximation and details;    reconstruct EBT approximation and details;    make table with SKT, EBT, associated approximation and details, and TSV; END;