Hybrid water wave optimization and support vector machine to improve EMG signal classification for neurogenic disorders

摘要

Electromyography (EMG) signal measures muscle response or electrical activity in response to nerve's simulation of muscle that is used for neuromuscular disorders diagnosis. This paper introduces a novel classification approach that hybridizes Water Wave Optimization (WWO) and Support Vector Machine (SVM) to improve EMG signal classification accuracy. Discrete Wavelet Transform (DWT) has been utilized for EMG noise cancellation and feature extraction, then SVM classifier was enhanced by WWO for the purpose of obtaining optimized parameters of SVM as well as finding optimal EMG features subset. The WWO-SVM yielded an overall accuracy of accuracy; 91.44% against 72% for the SVM classifier with RBF.