Classification of EEG Signals Using a Neural Network Model and Comparison with MLP and LSTM Algorithms

This study introduces a neural network model designed for the classification of EEG signals. The model, developed using the Python programming language, is trained with the backpropagation algorithm and successfully predicts the correct outputs of signals. The research demonstrates that the model effectively predicts actions by training on EEG signal features recorded during hand clenching and unclenching actions. The accuracy of the developed neural network model was compared with Multilayer Perceptron (MLP), widely used in biological signal classification, and Long Short-Term Memory (LSTM), a low-error learning algorithm. The obtained results are presented in the Performance Analysis section.