EEG Signal Classification using K-Nearest Neighbor Method to Measure Impulsivity Level

Abstract

Impulsivity is the tendency to act without considering consequences or without careful planning. It involves a quick response to a stimulus without sufficient consideration of the consequences. Impulsivity needs to be measured and detected because it has a significant impact on various aspects of a person's life. The factors that influence the level of impulsivity include social environment, stress level, mental health, and genetic factors. Impulsivity can be divided into multiple components, such as reduced sensitivity to unfavorable behavioral outcomes, a disregard for long-term implications, and quick and spontaneous responses to stimuli. Electroencephalogram (EEG) studies can identify specific brain wave patterns such as, Alpha, Betha, Theta, and Gamma waves everything based on an individual brain's level of impulsivity. Signals from the brain are processed to extract specific features that reflect the user's intentions. EEG records brain activity without surgery, and this information is used for the diagnosis, monitoring, and treatment of neurological diseases, as well as scientific research on the brain and mind. K-Nearest Neighbor (KNN) is a classification algorithm that functions by utilizing several K nearest data values (its neighbors) as a reference to determine the class of new data. The K-Nearest Neighbors (KNN) algorithm is used for classification, clustering, and pattern recognition in EEG data where clustering is in 4 classifications (Impulsive, Not Impulsive, Potentially Impulsive, and Very Potentially Impulsive). This classification model shows high accuracy (Training Data: 94.7%, Testing: 91.3%, and Validation Data: 91.8%). This research shows that the KNN algorithm is effective for assessing the degree of impulsivity.

Keywords

• eeg signal
• k-nearest neighbors
• brainwaves
• classification