Research on the Models for Forecast of Tax Revenue of Wenzhou City

Abstract

Tax revenue is a vital economic indicator that reflects the level of economic development, and tax revenue forecasting plays an important role in financial budgeting. Previous studies have demonstrated various influencing factors of tax revenue and proposed many feasible ways to forecast tax revenue. However, it is acknowledged that tax revenue of different region might bear different relation to influencing factors. In this research, tax revenue forecast of Wenzhou City is studied based on multiple linear regression model and MLP neural network model. The data for this research are collected from the website of Bureau of Statistics of Wenzhou, compiled in the 2022 statistical yearbook of Wenzhou. With multiple linear regression model, it is discovered that Value-added of the primary industry, Value-added of the tertiary industry, Investment in fixed assets, Total retail sales of consumer goods are significant for prediction. Comparing the forecasting outcomes of the two methods, the MLP neural network model appears to have better goodness of fit.