EEG Authentication System Based on One- and Multi-Class Machine Learning Classifiers

Abstract

In the current Information Age, it is usual to access our personal and professional information, such as bank account data or private documents, in a telematic manner. To ensure the privacy of this information, user authentication systems should be accurately developed. In this work, we focus on biometric authentication, as it depends on the user’s inherent characteristics and, therefore, offers personalized authentication systems. Specifically, we propose an electrocardiogram (EEG)-based user authentication system by employing One-Class and Multi-Class Machine Learning classifiers. In this sense, the main novelty of this article is the introduction of Isolation Forest and Local Outlier Factor classifiers as new tools for user authentication and the investigation of their suitability with EEG data. Additionally, we identify the EEG channels and brainwaves with greater contribution to the authentication and compare them with the traditional dimensionality reduction techniques, Principal Component Analysis, and χ2 statistical test. In our final proposal, we elaborate on a hybrid system resistant to random forgery attacks using an Isolation Forest and a Random Forest classifiers, obtaining a final accuracy of 82.3%, a precision of 91.1% and a recall of 75.3%.

Authentication using c-VEP evoked in a mild-burdened cognitive task

“…For authentication, precision is a very important performance indicator (Hernández-Álvarez et al, 2022 ). In the view of precision, the authentication using c-VEP evoked in the MBCT performed very well.…”

Cryptographic Algorithm Designed by Extracting Brainwave Patterns

“…Therefore, the verification stage requires a one-to-one match to decide the results as a valid user or imposter. Currently, classification/deep learning methods are used to perform user verification [8]. In this way, the analyzed literature suggests the following limitations:  Database: Most of the works that perform brain fingerprint extraction, and, in equal measure, the design of EEG-based authentication systems, use databases, designed for other purposes such as emotion recognition or character determination [32][33][34].…”

“…Unlike traditional systems, biometrics are advantageous as the encryption key [6] is harder to compromise or duplicate [7]. However, a biometric system is vulnerable to a variety of attacks designed to undermine the integrity of the authentication process [8,9].…”

Analysis of Cyber-Intelligence Frameworks for AI Data Processing

“…Although this application will make it possible to delve deeper into purely technical aspects, it should be borne in mind that the Diamond Model and Cyberkill Chain offer a better understanding of the socio-political sphere and its role in the context of an attack, which can help and, in many cases, be decisive in dealing with a cyberattack. Alternative works in the current literature regarding AI and cybersecurity and cyber intelligence usually focus on the elaboration of user authentication protocols [46][47][48]; network situation awareness [49,50]; dangerous behaviour monitoring [51,52]; and abnormal traffic identification [53,54]. In all cases, the main goal of the proposed schemes is to predict Alternative works in the current literature regarding AI and cybersecurity and cyber intelligence usually focus on the elaboration of user authentication protocols [46][47][48]; network situation awareness [49,50]; dangerous behaviour monitoring [51,52]; and abnormal traffic identification [53,54].…”

“…Alternative works in the current literature regarding AI and cybersecurity and cyber intelligence usually focus on the elaboration of user authentication protocols [46][47][48]; network situation awareness [49,50]; dangerous behaviour monitoring [51,52]; and abnormal traffic identification [53,54]. In all cases, the main goal of the proposed schemes is to predict Alternative works in the current literature regarding AI and cybersecurity and cyber intelligence usually focus on the elaboration of user authentication protocols [46][47][48]; network situation awareness [49,50]; dangerous behaviour monitoring [51,52]; and abnormal traffic identification [53,54]. In all cases, the main goal of the proposed schemes is to predict or identify an abnormal situation.…”