EEG Classification by Factoring in Sensor Spatial Configuration

Electroencephalography (EEG) serves as an effective diagnostic tool for mental disorders and neurological abnormalities. Enhanced analysis classification of EEG signals can help improve performance in classifying the A new approach is examined here enhancing using a novel model data representation that leverages knowledge spatial layout sensors. An investigation proposed provides evidence consistently higher accuracy compared with ignores sensor layout. The assessed models represent information content two different ways: one-dimensional concatenation channels frequency bands image-like two-dimensional channel locations. are used conjunction machine learning techniques. Performance these on tasks: social anxiety disorder classification, emotion recognition dataset, DEAP, physiological signals. We hypothesize will significantly outperform this validated our results yields 5-8% all algorithms investigated. Among investigated, Convolutional Neural Networks provide best performance, far exceeding Support Vector Machine k-Nearest Neighbors algorithms.