The Study Of The Influence And The Prediction Of Tax Based On Model Selection And Grey Analysis

Taxation is an important means of macro-economic control in a country,so the analysis of the influencing factors of tax revenue is indispensable in the study of the economic field.The analysis and prediction of the influence factors of tax revenue are crucial to the scientific decision-making and management of relevant departments.In fact,the influence factors of tax revenue are numerous and numerous,resulting in high cost and poor effect of management and decision-making,which should be answered both theoretically and empirically.Therefore,based on the model selection of point of view,using the LASSO method for high impact variables,and the joint in the theory of grey prediction GM(1,1)model established tax revenue factors affecting the combination of filtering and prediction model.This paper studied the research theory and methods of the related literature at home and abroad,this paper introduces the model select the relevant theory and grey system theory,including the subset selection method,method of NNG,LASSO and gray sequence generation,grey model,grey correlation analysis,grey prediction and so on.Through comparative analysis of the characteristics,advantages and connections of these two theories,the feasibility of establishing a combination model based on LASSO and grey theory is demonstrated.Secondly this article applies the theory of minimum variance of model prediction error,the LASSO model and GM(1,1)model which gives different combination weight coefficient,thus obtains the optimal combination forecast,as more as possible using different sample forecast information.The LASSO method and grey theory areis simulated numerically,and the simulation results show their superiority.Then based on the data of chongqing tax revenue from 1996 to 2015,the empirical analysis is carried out.The feasibility and superiority of the combined model are illustrated by comparing the performance of single LASSO method,single grey method and combination model in the example.Finally,this paper summarizes the analysis of the combination of the proposed model,and according to the performance of the portfolio model of outlook,can be used as part of the relevant departments for scientific decision and management for reference.