Identification of the Human Oculo-Motor System Based on the Volterra Series: Application in the Information Security System

The information technology of biometric identification a person has re-ceived further development due to the use as source primary data infor-mation models oculo-motor system (OMS) «input-output» type based on Volterra series. Eye-tracking is used build models. Experimental studies OMS two respondents were carried out. Based obtained with Tobii Pro TX300 eye-tracker, transient func-tions first, second and third orders when applying Volter-ra series model determined. This makes it possible increase accura-cy modeling and, result, reliability recognition in space proposed heuristic features, which are determined using inte-gral differential transformations multidimensional functions OMS, greatly simplifies features practical implementation Bayesian classifier.A high variability or-ders for was revealed. Thus, seems appropriate identification.A set proposed, basis from eye-tracking data. informativeness individual their combinations pairswas investigated. Two-dimensional feature spaces maximum value probability correct indicator solving problem found (Pmax=0.974). research results construction clas-sifiers different by means machine learning formed trainingsamples