A novel feature subset selection algorithm using artificial bee colony in keystroke dynamics

Abstract

Keystroke dynamics is a biometric technique to identify a user based on the analysis of his/her typing rhythm. The steps of keystroke dynamics include feature extraction, feature subset selection and classifier. In the experiment mean, median and standard deviation of feature values such as latency, duration and digraph are measured in feature extraction. A meta-heuristic based on artificial bee colony algorithm is proposed for feature selection. Neural network is used for classification. The performance of artificial bee colony algorithm is analyzed with regard to feature reduction rate. The obtained result shows comparable quality with faster convergence. © 2012 Springer India Pvt. Ltd.