A3 Refereed book chapter or chapter in a compilation book
EEG-based Biometric Authentication Using Wavelet Packet Decomposition and Ensemble Classifiers
Authors: Subasi, Abdulhamit; Subasi, Muhammed Enes; Qaisar, Saeed Mian
Editors: Naik, Ganesh R.
Edition: 1st Edition
Publisher: CRC Press
Publication year: 2026
Book title : Time-Frequency Analysis in Biomedical Engineering : Contemporary Methods and Applications
First page : 244
Last page: 265
ISBN: 978-1-041-00745-6
eISBN: 978-1-003-61144-8
DOI: https://doi.org/10.1201/9781003611448-17
Publication's open availability at the time of reporting: No Open Access
Publication channel's open availability : No Open Access publication channel
Web address : https://doi.org/10.1201/9781003611448-17
Abstract
In the digital transformation era, the need for secure and reliable authentication methods has become paramount, particularly in applications involving sensitive information access and identity verification. Biometric authentication, which relies on individuals’ unique physiological or behavioral traits, offers a promising solution to address these security challenges. Electroencephalography, a non-invasive technique for recording electrical activity in the brain, has emerged as a potential biometric modality due to its inherent uniqueness and stability. This chapter explores the application of wavelet packet decomposition (WPD) to EEG-based biometric authentication, aiming to enhance the authentication system’s accuracy, efficiency, and robustness. WPD, a powerful signal processing technique, provides a time–frequency representation of EEG signals, enabling the extraction of discriminative features for authentication purposes. The proposed methodology involves several key stages. First, WPD decomposes the EEG signals into different frequency bands, capturing both local and global signal characteristics. Statistical feature extraction techniques are then employed to extract relevant biometric features from the WPD coefficients, which serve as inputs to the authentication model. Subsequently, ensemble machine learning models are utilized for biometric authentication. To evaluate the effectiveness of the proposed approach, extensive experiments are conducted using the BCI200 EEG dataset. Performance metrics such as accuracy, F-measure, area under the curve, and Kappa are utilized to assess the authentication system’s efficacy in accurately verifying individuals’ identities. The results demonstrate that the combination of WPD with ensemble machine learning models for EEG-based biometric authentication yields promising outcomes, with high accuracy and low error rates achieved across various scenarios. Moreover, the proposed methodology exhibits robustness to noise and variability in EEG signals, indicating its potential for real-world deployment in security-sensitive applications.
In the digital transformation era, the need for secure and reliable authentication methods has become paramount, particularly in applications involving sensitive information access and identity verification. Biometric authentication, which relies on individuals’ unique physiological or behavioral traits, offers a promising solution to address these security challenges. Electroencephalography, a non-invasive technique for recording electrical activity in the brain, has emerged as a potential biometric modality due to its inherent uniqueness and stability. This chapter explores the application of wavelet packet decomposition (WPD) to EEG-based biometric authentication, aiming to enhance the authentication system’s accuracy, efficiency, and robustness. WPD, a powerful signal processing technique, provides a time–frequency representation of EEG signals, enabling the extraction of discriminative features for authentication purposes. The proposed methodology involves several key stages. First, WPD decomposes the EEG signals into different frequency bands, capturing both local and global signal characteristics. Statistical feature extraction techniques are then employed to extract relevant biometric features from the WPD coefficients, which serve as inputs to the authentication model. Subsequently, ensemble machine learning models are utilized for biometric authentication. To evaluate the effectiveness of the proposed approach, extensive experiments are conducted using the BCI200 EEG dataset. Performance metrics such as accuracy, F-measure, area under the curve, and Kappa are utilized to assess the authentication system’s efficacy in accurately verifying individuals’ identities. The results demonstrate that the combination of WPD with ensemble machine learning models for EEG-based biometric authentication yields promising outcomes, with high accuracy and low error rates achieved across various scenarios. Moreover, the proposed methodology exhibits robustness to noise and variability in EEG signals, indicating its potential for real-world deployment in security-sensitive applications.
UTU Research Portal