ABSTRACT

Motor image Critical disease diagnosis relies heavily on EEG classification. The complexity of the motor imagery EEG data hindered accurate classification. The motor imagery EEG classification rate is increased using the feature Optimization procedure. A deep neural network-based classifier for motor imagery EEG classification was proposed in this paper. The design deep neural network is a three-layer neural network model that incorporates the teacher learning-based optimization and feature optimization technique. The EEG data's noise and artefacts are reduced by a teacher learning-based optimization technique, which also enhances the input vectors for DNN. The suggested algorithm has been tested on datasets from the third and fourth BCI competitions and has been simulated in MATLAB environments. According to the evaluation's findings, the suggested algorithm compresses the current motor imagery EEG categorization technique quite effectively.

Keywords: : Motor Imagery (MI) EEG, Bayesian Feature Extraction, TLBO, DNN, Wavelet Transform, MATLAB, DWT, Optimization, BCI