Back propagation neural network (BPNN) as tracing method to trace physionet EMG signals: a case study

摘要

A new Modified Levenberg–Marquardt Algorithm (M-LMA) different fromLevenberg–Marquardt Algorithm (LMA) was used to trace the Physionet EMGsignals in tanning back propagation neural network (BPNN). M-LMA and LMAwere simultaneously used to minimize back propagation errors in tanning BPNNto trace the Physionet EMG signals under the same learning rates of 0.1. Resultedshown M-LMA was better then LMA in training BPNN and could be as a bettertracing method in this case.