Assessment of Human Random Number Generation for Biometric Verification

Abstract

 

Random number generation is one of the human abilities. It is proven that the sequence of random numbers generated by people do not follow full randomness criteria. These numbers produced by brain activity seem to be completely non stationary. In this paper, we show that there is a distinction between the random numbers generated by different people who provides the discrimination capability, and can be used as a biometric signature. We considered these numbers as a signal, and their complexity for various time-frequency sections was calculated. Then with a proper structure of a support vector machine ( SVM ), we classify the features. The error rate obtained in this study, shows high discrimination capabilities for this biometric characteristic.