A Novel Non-Linear Transformation Based Multi User Identification Algorithm for Fixed Text Keystroke Behavioral Dynamics

In this paper, we propose a new technique to uniquely classify and identify multiple users accessing single application using keystroke dynamics. This problem is usually encountered when have legitimate access shared computers accounts, where, at times, one user can inadvertently be logged in on another user’s account. Since the login processes are bypassed stage, rely dynamics order tell apart. Our algorithm uses quantile transform techniques from localization users. Specifically, use an known as ordinal Unfolding based Localization (UNLOC), which only data obtained comparing distance proxies, by “locating” reduced PCA/Kernel-PCA/t-SNE space their typing patterns. results validated with help of benchmark datasets show that our outperforms other methods.