Design of biometric system and modeling of machine learning for entering the information system
Mariya Nazarkevych, Nataliia Lotoshynska, Volodymyr Hrytsyk, Bohdana Havrysh, Olha Vozna, Oleksandra Palamarchuk · 2021 IEEE 16th International Conference on Computer Sciences and Information Technologies (CSIT) · 2021
Today on the market there is a stable and significant increase in demand for biometric access control systems, fingerprints are on the priority lists. Approaches to the recognition of the biometric system of fingerprints based on machine learning methods have been developed. The system performance evaluation determines two parameters, False Acceptance Rate, and False Rejection Rate. In order to meet the requirements for fingerprint verification systems, which are FAR = 10−4and FRR = 10−2, the method of machine learning with the teacher was used. Experimental results were obtained, which showed the effectiveness of the proposed method.