Biometric authentication and identification using behavioral biometrics technique of signature verification

Abstract

Authentication using biometrics is on the rise due to various security concerns and especially in India with the advent of Digital India concept promoting the wide use of aadhar as an authentication mechanism. The authentication using a login and a password are fading away due to the various malpractices and fraudulent methods that have been developed which pose a threat to this kind of authentication and security mechanism. Signature verification is one of the important aspects of biometric authentication which is slowly finding its use in certain niche organizations. The present study uses an ensemble approach, stacking to authenticate or validate a user by his signature and proposes a hybrid model using three algorithms namely, Random Tree, Logistic Regression & Multi-Layer Perceptron for achieving better accuracy.