Analysis And Short-term Forecast Of Influencing Factors Of Chengdu’s Financial Revenue

Local revenue is an important indicator of the financial strength of a regional government,and the scope and quantity of goods and services that the government needs to provide in various economic activities(public expenditure)depends to a considerable extent on the government’s fiscal revenue.The current fiscal revenue of Chengdu is on a stable and increasing trend.Now,with the realization of the "Chengdu-Chongqing Urban Agglomeration Development Plan",Chengdu City will have greater development opportunities in the future,so there is a greater need for scientific and reasonable revenue forecasting in order to improve the management ability,so that the revenue and expenditure in the budget will no longer be arbitrary and blind,and the coordinated development of all aspects of the local community can be promoted.Therefore,the establishment of a more accurate revenue forecasting mechanism is of positive significance for the future development and social stability of Chengdu.In this paper,the indicators affecting Chengdu city are initially selected based on the previous research in related literature.Since we found that the Chengdu City Statistical Yearbook was only updated to the year 2020 during the completion process,19 relevant data and related indicators of Chengdu City from 1994 to 2020 were selected.Firstly,Lasso regression model and gray correlation analysis were used to select the influencing factors respectively,and it was found that the results obtained from Lasso regression model were not consistent with the qualitative analysis.In contrast,the gray correlation analysis was used to derive the correlation degree of each indicator and rank the correlation degree so as to filter out 9 influencing factors that affect the fiscal revenue of Chengdu,and these 9influencing factors are consistent with the qualitative results.We consider using the gray prediction model for the prediction of each influencing factor to prepare for the content of fiscal revenue later.Given that the gray forecasting model is suitable for situations where the amount of data is small,we selected sample data from the most recent years to predict the forecast values of the 9 influencing factors for 2021 and 2022,and subsequently used a gray Markov model to correct the forecast results.For the prediction of fiscal revenue,the real values of the nine influencing factors were substituted into the support vector machine regression model and the BP neural network model for prediction,and the best prediction model was selected by comparing the accuracy and model fit of the test samples,and the results showed that the BP neural network model had better results.Finally,we substitute the predicted values of the nine influencing factors in 2021 and 2022 into the trained BP neural network model to predict the revenue data of Chengdu in 2021 and 2022,which shows that the revenue of Chengdu shows a continuous increasing trend.Then we substituted the predicted values of nine influencing factors in 2021 and 2022 into the trained BP neural network model to predict the revenue data of Chengdu in 2021 and 2022,which showed that the revenue of Chengdu showed a continuous rising trend.Finally,with the existing policy background and related analysis results,we propose some guidance for the future highquality development of Chengdu city’s fiscal revenue.