Short-Term and Mid-Term Population Projections Especially of Germany via WASD Neural Network: Theoretical Parts

Population has a direct impact on the demands of consumption, resource, and employment. Population factors, such as size and growth rate, are crucial in making decisions and plans of trade, environment protection, resource allocation and human resource development. Thus, population projection is an essential part of strategy development for companies as well as governments, and draws the attention of many researchers. Some projection methods have been developed, but most of the popular ones require a vast knowledge in demography and only can be operated by skilled professionals. In this paper, as an improvement, we theoretically develop a regression model and a subtrend model of WASD (weights and structure determination) neural network in a simultaneous manner to make population projections especially of Germany in the midterm and near future. Specifically, we aim at utilizing the learning capabilities of the WASD neural-network models with single input and multiple inputs, having been very popular recently, to propose these easy-used and high accuracy models for verifications and projections via later computer experiments.