Authenticating computer access based on keystroke dynamics using a probabilistic neural network

Abstract

Most computer systems are secured using a login id and password. When computers are connected to the internet, they become more vulnerable as more machines are available to attack them. In this paper, we present a novel method for protecting/enhancing login protection that can reduce the potential threat of internet connected computers. Our method is based on and enhancement to login id/password based on keystroke dynamics. We employ a novel authentication algorithm based on a probabilistic neural network. Our results indicate that we can achieve an equal error rate of less than 5%, comparable to what is achieved with hardware based solutions such as fingerprint scanners and facial recognition systems.