A Wrapper-Based Classification Approach for Personal Identification through Keystroke Dynamics Using Soft Computing Techniques

Abstract

The password is the most widely used identity verification method in computer security domain. However, due to its simplicity, it is vulnerable to imposters. A way to strengthen the password is to combine Biometric technology with password. Keystroke dynamics is one of the behavioural biometric approaches which is cheaper and does not require any sophisticated hardware other than the keyboard. The chapter uses a new feature called Virtual Key Force along with the commonly extracted timing features. Features are normalized using Z-Score method. For feature subset selection, Particle Swarm Optimization wrapped with Extreme Learning Machine is proposed. Classification is done with wrapper based PSO-ELM approach. The proposed methodology is tested with publically available benchmark dataset and real time dataset. The proposed method yields the average accuracy of 97.92% and takes less training and testing time when compared with the traditional Back Propagation Neural Network.