Utilization of Levenberg-Marquardt based Neural Network Classifier in EMG signal Classification

Abstract— Electromyography (EMG) signal provides a significant source of information for identification of neuromuscular disorders. This paper presents an application of neural network classifier on classification and identification of different normal and auto aggressive actions of hands and legs. Eight features that are extracted from eight channel EMG signals representing these actions have been used as inputs to the artificial neural network classifier. Levenberg – Marquardt algorithm is used as the learning algorithm of the neural network classifier in the present work. This paper also presents a comparison between the performances of Levenberg – Marquardt based neural network classifier and Support Vector Machine Classifier (SVM) in classification of EMG signal.