Vykhovanets Y. Neural network models of estimation of the functional states of human

The purpose: working out of models of a quantitative estimation of functional conditions of the person on the basis of neural networks. Methods of researches: anthopometrical (measurement of weight and length of a body), physiological (measurements of arterial pressure and frequency of warm reductions), variabilities of a warm rhythm, stabylometry indicators), psychophysiological (speed visually-momtornyh reactions), psychological, clinical (survey), mathematical (statistical methods, neural network models). The received results: on indicators stabylometry the nonlinear five-factorial forecasting model of functional conditions of the person is developed. The model of a quantitative estimation of functional conditions on the basis of calculation of predicted value of biological age is developed. Critical value of risk of deterioration of functional condition Ykr is received., it allows to divide surveyed on two groups: with high and low risk of functional conditions. Are developed neural network models of an estimation of functional conditions. Efficiency: increase in relative density recovered in 2 times, disease decrease for the first time the revealed diseases on 20,3 %, improvement of quality of life in 1,5 times. Introduction areas: scientific research institutes and psyhophisiology laboratories, university clinics, treatment-and-prophylactic establishments, chairs of physiology, therapy, neurology, psychiatry.