Understanding the Impact of EEG Segment Length on Biometric Authentication

Biometric authentication becomes more essential as technology evolves. The ease with which a biometric can be used affects its impact and industrial implementation. One promising biometric is electroencephalography (EEG). This paper investigates how to optimize identification accuracy in EEG-based biometric authentication without overly complicating or slowing down the system. To determine the point at which the EEG data is most useful for authentication, we examine various segment lengths of the data and apply machine learning models. Our research will contribute to the advancement of EEG biometric technologies by offering a safe and convenient identity verification solution. The research shows that accuracy increases by increasing the EEG segment length until reaching a turning point and then continuing with a plateau. The turning point was statistically determined to be 2 seconds. This finding contributes to the possibility of applying EEG-based biometric authentication in uncontrolled real-world settings.