ABSTRACT
This chapter introduces the basic theories and developments of the neuronet with back-propagation (BP) and weights-and-structure-determination (WASD) algorithms. It explores the power-activation neuronet models with WASD and BP algorithms are applied to predicting Russian population based on the 2013-year historical population data, respectively. The power-activation neuronet with WASD algorithm is illustrated to be a better one than the power-activation neuronet with BP algorithm for Russian population prediction. Russian population problem attracts great concerns to the future trend of population and the development of the nation. As the conventional BP neuronet has some inherent weaknesses, two types of improved feed-forward neuronet are constructed for the Russian population prediction. More specifically, a type of three-layer power-activation neuronet equipped with the BP algorithm and a type of three-layer power-activation neuronet equipped with the WASD algorithm are built on the basis of the 2013-year historical population data for the Russian population prediction.
