DESIGN AND EVALUATION OF AN SVM-BASED BIOMETRIC AUTHENTICATION SYSTEM USING KEYSTROKE DYNAMICS
Ключевые слова:
Biometric authentication, keystroke dynamics, Support Vector Machines (SVM), user identification, machine learning, FRR, FAR, security, latency, pressure force.Аннотация
This study considers the issues of creating a biometric authentication system based on the dynamics of keystrokes and improving its efficiency. Since traditional password-based authentication methods are not sufficiently reliable from a security perspective, a biometric authentication method based on the analysis of users' keyboard typing habits is proposed. The study uses the Support Vector Machines (SVM) model to analyze and classify user typing patterns. An authentication system is developed based on keystroke force and latency data, and its False Rejection Rate (FRR) and False Acceptance Rate (FAR) indicators are evaluated. The results obtained demonstrate the efficiency of the authentication system based on the dynamics of keystrokes using the SVM model. The results of the study pave the way for exploring the possibilities of integrating keyboard biometrics with other authentication methods in the future.
