江西省地方财政收入预测与研究

Forecast and Research on Local Fiscal Revenue of Jiangxi Province

摘要: 财政收入是国家建设和治理的基础和重要保障，它不仅能够综合反映出国民经济状况，也为国家政府对市场经济进行宏观调控奠定了基础。本文以江西省财政收入为例，收集当地近20年的相关数据进行实证分析。首先利用LASSO回归分析筛选出影响江西省财政收入的重要因子，将其作为预测模型的变量。然后将灰色GM(1,1)与BP神经网络结合起来，先利用GM(1,1)拟合得到2025年各变量的估计值，再将这些估计值作为BP神经网络的输入，输出值即为江西省2025年的财政收入预测值。实验结果表明，预测模型精度较高，可为当地政府制定相关财政政策给出理论参考。

Abstract: Fiscal revenue is the basis and important guarantee for national construction and governance. It can not only comprehensively reflect the state of the national economy, but also lay the foundation for the national government to carry out macro-control over the market economy. This paper takes the fiscal revenue of Jiangxi Province as an example and collects the relevant data in the past 20 years for empirical analysis. First of all, LASSO regression analysis was used to screen out the important factors affecting the fiscal revenue of Jiangxi Province and take them as the variables of the prediction model. Then the grey GM(1,1) is combined with BP neural network, and the estimated value of each variable in 2025 is obtained by the fitting of GM(1,1). Then these estimated values are used as the input of BP neural network, and the output value is the predicted fiscal revenue value of Jiangxi Province in 2025. The experimental results show that the prediction model has high accuracy and can provide theoretical reference for local government to formulate relevant fiscal policies.