Comparative Study of ANN Algorithms for EMG Signals

Electromyography (EMG) signals play a pivotal role in medical fields, rehabilitation robotic and human–computer interaction. Machine learning tools help to decompose and process EMG signals acquired from the muscles. The purpose of this study is to analyze artificial neural networking (ANN) algorithms for EMG signals in order to obtain efficient and effective ways to understand signals. Two algorithms of ANN are compared, which are Levenburg– Marquardt and scale conjugate gradient algorithms on the basis of their accuracy in terms of regression values. It was found that both algorithms showed efficient results but Levenburg–Marquardt gave better results for given sets of EMG signals.