Use of RBF neural network in EMG signal noise removal
WSEAS Transactions on Circuits and Systems
| Hi-index | 0.00 |
Filtering of signals is of primary importance in signal processing. The design of filters to perform signal estimation is a problem that freeze up in the design of communication systems, control systems, in geophysics & in many other applications & disciplines. Optimum filters are used for linear prediction. In this paper, neural networks have been trained to predict a signal using the past signal samples. It is found that neural networks such as multiplayer perceptron, general feed forward, modular neural network, etc., comprising of three hidden layers with a linear transfer function elegantly filters various signals under consideration. Keywords Linear Estimation, Newlind, Feed-forward Neural Network, RLS, MSE, MLP, RBF.