Keystroke Dynamic Classification using Machine Learning for Password Authorization
Yohan Muliono, Hanry Ham, Dion Darmawan · Procedia Computer Science · 2018
Many methods used to perform a password authentication using user’s biometrics such as fingerprint recognition, retina recognition, voice recognition, etc. However, additional sensors needed to perform most of biometric recognition methods and will be invasive to users caused by additional tools needed to perform a password authentication. Keyboard Dynamics is one of the solution to perform password authentication without adding any tools which being disruptive to some users. The biometric keystroke dynamic system is relatively unexplored compared to other behavioral authentications discipline. Coupled with the limited number of studies that have been done compared with other biometric systems. Several machine learning research has been conducted but few of them applying deep learning for solving this problem. This research will be focusing in deep learning using optimizer to beat the previous research which using another machine learning techniques. This research shows a better result using optimizer in deep learning resulting in 92.60% accuracy.