Support vector machine analysis in mouse dynamic authentication classification

Authentication systems have built the foundation for validating and securing user identities. One such authentication method is Mouse Dynamics Authentication. Mouse Dynamics Authentication is used to identify users based on mouse movements on a system. Everyone’s mouse movement behavior is unique, and this unique trait is the basis of its security. The more advanced technology is followed by the human need for the security of personal data is higher due to the increasing ability of attackers to steal data. To improve the security of personal data, it is necessary to reduce the FAR and FRR value in this research. The authors propose combining the Support Vector Machine method and Simple Random Sampling as dataset sampling. The results of this research indicate that the proposed SVM method provides good results. This research provides an opportunity to be implemented in a real login system because our method gives the best results with a False Acceptance Rate (FAR) of 0.0457 and a False Rejected Rate (FRR) of 0.0613.