我国教育水平影响因素研究

Research on Influencing Factors of China’s Educational Level

摘要: 以2015年到2019年31个省份的人均受教育年限为研究对象，从人口、政策、经济、科技水平、服务供给五个大方面选择了10个影响因素。首先运用弹性网对影响教育水平的因素进行筛选压缩，并选用岭回归模型、Lasso回归模型作为对比。最终得出较为合适的变量进行参数估计，最后对模型的准确率进行预测。得出最终影响因素分析结果。

Abstract: Taking the per capita years of schooling in 31 provinces from 2015 to 2019 as the research object, 10 influencing factors were selected from five aspects: population, policy, economy, scientific and technological level, and service supply. Firstly, elastic net was used to select and compress the factors affecting education level, and Ridge regression model and Lasso regression model were selected as comparison. Finally, more appropriate variables are obtained for parameter estimation, and finally the accuracy of the model is predicted. The final analysis results of influencing factors are obtained.