Online Signature Verification Using Logistic Regression

Online signatures are one of the widely accepted biometrics used for the purpose of authentication and identification. Although there are several approaches and algorithms used in online signature verification systems, nowadays machine learning algorithms are leading the field. In this paper, an online signature verification system based on logistic regression is presented. The SVC2004 database was used to test the accuracy of the system by applying different combinations of features. The results showed that we could achieve accuracies between 91.7% and 98.08%.