Metric based technique in multi-factor authentication system with artificial intelligence technologies

Abstract

The metric based authentication system have found a lot of applications in systems such as the banking and immigration point based systems used in several countries such the United Kingdom, Australia and Canada. Scoring based authentication system is also widely used in biometrics authentication systems. In this paper, we extend the metric based authentication system to come up with a web based multi-factor authentication system. The system uses four identity attributes with artificial neural networks (ANN) used to fuse the metrics values in an information fusion system. Levenberg-Marquardt Optimization is used to update the neural weights and bias while trainln is used to train the ANN information fusion system. Delaunay triangulation technique is used for the fingerprint detection algorithm. © 2012 Springer-Verlag GmbH.