A population prediction model based on generalized regression neural network
基于广义回归神经网络的人口预测模型

摘要

人口问题一直是社会发展问题的核心，是制约社会向前发展的关键因素，人口增长的过快或过慢，都会对世界发展带来一定的损失，正确的引导是解决人口问题的重要手段，为建立人口预测的数学模型，分别预测2100年底中国、印度和全球的总人口数，并分析人口的变化趋势。通过建立广义回归神经网络模型，使用Matlab进行编程预测，分别对1950~2021年的中国、印度和全球的总人口数进行预测，并与实际人口数量进行比较，证明模型的可行性。再进一步预测2022~2100年中国、印度和全球的人口总数，对结果进行可视化处理。

Abstract

Population problem has always been the core of social development and a key factor restricting social development. Too fast or too slow population growth will bring certain losses to world development. Correct guidance is an important means to solve the population problem. In order to establish the mathematical model of population prediction, the total population of China, India and the world at the end of 2100 is predicted respectively, and the changing trend of population is analyzed.By establishing a generalized regression neural network model and using Matlab for programming prediction, the total population of China, India and the whole world from 1950 to 2021 are predicted respectively, and the actual population is compared to prove the feasibility of the model. Futhermore, the total population of China, India and the world in 2022-2100 is predicted, and the results are visualized.