RISK-ADAPTED ACCESS CONTROL WITH
MULTIMODAL BIOMETRIC IDENTIFICATION
Gábor Werner1 and Received: 31st January 2019. ABSTRACT KEY WORDS CLASSIFICATION
1Óbuda University, Applied Biometric Institute
2Óbuda University, Institute of Mechatronics and Autotechnics
INDECS 18(3), 327-336, 2020
Full text available here.
Accepted: 3rd March 2020.
The presented article examines the background of biometric identification. As a technical method of authentication, biometrics suffers from some limitations. These limitations are due to human nature, because skin, appearance and behavior changes more or less continuously in time. Changing patterns affect quality and always pose a significantly higher risk. This study investigated risk adaption and the integration of the mathematical representation of this risk into the whole authentication process. Several biometrical identification methods have been compared in order to find an algorithm of a multimodal biometric identification process as a possible solution to simultaneously improve the rates of failed acceptations and rejections. This unique solution is based on the Adaptive Neuro-Fuzzy Inference System and the Bayesian Theorem.
multimodal biometrics, artificial intelligence, ANFIS, risk management
Received: 31st January 2019.