A Comparative Analysis of EEG Signals Using PSD, FFT and Wavelet Transform Methods

The study of the human brain heavily relies on electroencephalography (EEG) in today’s world, providing crucial insights into its functioning. Despite its importance, the classification of EEG signals poses a significant challenge due to their complex and dynamic nature. This paper aims to improve the accuracy of classifying EEG signals related to motor imagery (MI) by utilizing robust techniques such as power spectral density (PSD), fast fourier transforms (FFT), and wavelet transforms (WT). These methods, widely recognized and utilized in both academic and industrial settings, are combined with machine learning algorithms including K-Nearest Neighbours (KNN), Support Vector Machine (SVM), Quadratic Discriminant Analysis (QDA), Linear Discriminant Analysis (LDA), and Random Forest Algorithm (RFA) to enhance the classification accuracy. This holistic approach aims to address the aforementioned challenges. The goal is to help advance the classification of EEG signals, which will in turn make it easier to explore and understand the dynamics of the brain.