Authenticating users in real world applications using multimodal biometric system for smartphones based on support vector machine compared with decision tree

• Main Authors: Sandeepkumaryadav, G., Vinodhini, G. Arul Freeda
• Format: Conference Proceeding
• Language: English
• Subjects: Biometrics; Datasets; Decision trees; Electronic devices; Smartphones; Support vector machines
• ISSN: 0094-243X; 1551-7616
• DOI: 10.1063/5.0232869

The objective of this project is to improve the authentication of users in applications that are used in the real world by utilising multimodal biometric systems for mobile devices. This will be accomplished through the usage of mobile devices. The Constituents and the Methods involved: This investigation made use of a research dataset that was initially sourced from the Kaggle database system. This dataset was used for the investigation. SVM and DT carried out a number of testing and training divisions in order to accomplish the goal of authenticating users in real-world applications that make use of multimodal biometric systems for smart phones. When doing the Gpower test, it was seen that about 85 percent was utilised (the configuration settings for Gpower were α=0.05 and power=0.85). On the basis of the findings, it is possible to draw the conclusion that Support Vector Machines (SVM) exhibit a greater level of accuracy compared to Decision Trees (DT), with a significance level of 0.001 (p