Power-Activated WASD Neuronet Based Russian Population Estimation, Correction, and Prediction

In recent years, the serious situation of Russian population causes concerns of the government, which also attracts great attentions of researchers from all over the world. The research on Russian population can help the government to make positive policies for solving the crisis and boosting economic growth. In this paper, implicitly considering almost all factors that influence population development, we present a 3-layer feed forward power-activated neuronet (PAN) equipped with the weights-and-structure-determination (WASD) algorithm for the estimation, correction and prediction of Russian population. Many numerical tests are conducted via WASD-PAN using past 2013-year Russian population data. We estimate the Russian population from 1000AD to 1800AD, correct it around 1897AD, and further indicate several possibilities of Russian population in the future. With the most possibility, Russian population is predicted to decrease steadily in next decade, while it is still possible that Russian population will (finally) increase.