Rules-Based and SVM-Q Methods With Multitapers and Convolution for Sleep EEG Stages Classification

Sleep EEG signals analysis is an approach that helps researchers identify and understand the different phenomena concealed within sleep data. This research introduces a time-frequency to untangle parameters of stages classification from computes spectral estimation signal based on set controlled wavelets using multitaper with convolution (MT&C) method. In this study, MT&C methods implemented extract features single data channel. Then two separated approaches are applied for stage classification. The first one waves characteristic definitions (named as Rules-based method) directly classify each 30 second segment after feature extraction. uses support vector machine quadratic equation (SVM-Q) classifier experts’ scoring. experimental results evaluated, outcomes show overall accuracy 90% average sensitivity 96.2% specificity 93.2% SVM-Q 87.6% method healthy subjects. On other hand, subjects abnormal 78.1% 73.4%