Accuracy comparison of classification techniques for mouse dynamics-based biometric carp

Abstract

Combining more than one authentication schemes enhances the robustness of a system against cyber-attacks. In this paper, we introduce a novel mouse dynamics-based biometric CaRP (CAPTCHA as gRaphical Password). It combines mouse dynamics-based biometric authentication scheme with knowledge-based authentication scheme. This study primarily focuses on the comparison of classification accuracy of binary decision tree, SVM, and ANN for proposed mouse dynamics-based authentication scheme of CaRP.