Classification of normal and relaxed conditions based on brain signal activity with Electroencephalography using k-Nearest Neighbor (kNN)

Abstract

Research on how to think, patterns that occur when humans act, human conditions observed through brain waves, and various studies related to the human nervous system and coordination are still extensive in scope for development. Electroencephalography (EEG) is an instrument commonly used to develop research related to the mechanism of brain wave activity. Changes in the brain's electrical potential can be exploited by carrying out specific analyses using various signal processing methods, which are grouped into various categories of waves, including delta (0 - 4 Hz), theta (4 - 7 Hz), alpha (8 - 12 Hz) and beta (12 – 30 Hz). This research has succeeded in building a classification model through several stages, namely preprocessing with filtration and data extraction, data processing consisting of clustering using the K-Means algorithm, and classification using the k-Nearest Neighbor (kNN) algorithm to calculate the accuracy value of the model. The classification process produces two categories of conditions: normal and relaxed. The results of testing the classification model using the k-Nearest Neighbor (kNN) produce an accuracy value of 88%.