METHOD OF ACTIVITY EFFICIENCY PREDICTION FOR ALMOST HEALTHY INDIVIDUALS BASED ON SET OF ELECTROPHYSIOLOGICAL PARAMETERS

摘要

FIELD: medicine.;SUBSTANCE: invention refers to medicine, namely to labor physiology and health. The investigated persons are divided into groups with the known characteristics of simulated activity efficiency based on the analysis of the Gorbova-Schulte test results. During the test, electrophysiologic parameters are recorded: EEG, cognitive evoked potentials P300, electrocardiograms. At that, a network represented by a multilayer perceptron with 21 input neurons, 24 neurons in the intermediate layer and an output neuron is used as the neural net.;EFFECT: invention can improve the accuracy of prediction, which is achieved through application of the neural net and optimum complex of neurophysiological parameters.;1 dwg, 3 tbl