Возможности моделирования предрасположенности к наркозависимости методами искусственного интеллекта

Вестник Пермского университета. Философия. Психология. Социология. 2015. № 1. С. 61–71.

Yasnitsky L., Грацилев В. И., Куляшова Ю. С., Черепанов Ф. М.

A computer program designed to determine the degree human predisposition to drug addiction. The program is based neural network trained on the results of sociological surveys. Error of neural network model was less than 1%. With the help of neural network model evaluated the importance of factors that can influence the predisposition to drug addiction. The most important factors were: the level of education, having friends who use drugs, temperament type, number of children in the family, financial situation. Neural network model allows to evaluate the effect of varying the parameters characterizing the man and his predisposition to addiction and select the optimal combination of these parameters for each individual and thus receive individual recommendations for reducing drug addiction.

Keywords: нейронная сеть, прогноз, экономико-математическое моделирование, mathematical modeling, artificial neural networks, drug addiction, prediction, artificial intelligence