Research On Tax Forecasting Model And Its Application In The Rent Of GuiZhou Province

The tax revenue forecast is an important reference and basis for national economic decision-making and budgeting.Finding out the related factors on the impact of the tax's law through the analysis of a large number of the historical data,it is the establishment of tax analysis and forecast model.It has important significances for improving the accuracy and timeliness of prediction and analysis of tax.Because of the complexity of the tax component of the tax gross direct mining analysis of the accuracy of the prediction method is difficult to further improve.Especially for some specific areas,the major elements of the tax effect of the impact with the mainstream elements are different,the same as the main industry's tax they focus on.So the different characteristics of different regions should be used the different prediction methods.That is very necessary.This paper analyzes the common tax forecasting model of the forecast and characteristics of the forecast,and puts forward the combination forecasting model,the error of the model is the smallest and the practicality is the strongest.The scheme from Guizhou Province tax professional management platform of tax analysis module to obtain the original data,the data in Liupanshui City as an example,using SPSS and linear regression prediction model,curve regression forecast model and ARIMA model composed of combination forecasting model and other model so as to obtain the forecast result.According to different models to predict the tax revenue of the next year,according to the forecast data to estimate the tax capacity of the coming year,and make the corresponding plan adjustment.The results show that the combined forecasting model has good forecast effect.