EEG independent component classification using ensemble based technique

This Research delves into the effectiveness of convolutional neural networks (CNNs) in automating the classification of EEG-independent components, a critical task in neuroscientific research and clinical applications. Our research aims to comprehensively evaluate CNN models’ performance in accurately classifying EEG-independent components, considering diverse aspects, including activation functions,hyperparameters, and feature extraction techniques. Through an extensive investigation, we explore the impact of different data representations, such as time series, power spectral density, and topo plots, on classification accuracy. Furthermore, we incorporate decision-level fusion and CNN architecture modificationstoenhance classification performance. ThisResearch contributes to advancing the state-of-the-art in EEG signal analysis, paving the way for improved accuracy and efficiency in EEG artifact classification