Modeling of Multifactor Regression of the Synchronization Period for an Individual Pattern of the Human Brain Neural Network

Abstract

We consider multifactor model of the human brain synchronization depending on external factors (frequency and index of alpha rhythm, interhemispheric asymmetry and type of neural network of the head brain). All statistical data were obtained with the help of device Neuron-Spectrum (Neurosoft, Russia), which carried out multichannel registration of EEG (electroencephalogram) by 8 cup electrodes connected with ear electrodes and localized in accordance with the system 10-20. We consider the following two groups of subjects: people who regularly practice psycho-physical relaxation (meditation) and aged 22-34, and people who do not practice psycho-regulation and aged 22-38. The testing of the constructed models for the Control group proves an importance of all investigated factors and gives a mean absolute percentage error 0.53%. Also, the coefficient of determination showed that 99.9% of dispersion of the investigated feature is explained by the considered external factors. For the PPR (psycho-physical relaxation) group, we prove an importance of only one of the considered factors (frequency of alpha rhythm). Therefore, for this group, an additional research is needed to search for more significant external factors. Keywords: multifactor model; synchronization of the head brain; interhemispheric asymmetry; type of neural network of the head brain; frequency of alpha rhythm; index of alpha rhythm; modeling; electroencephalogram.