Effect Analysis Of Root Test Of Degenerative Unit Based On Non-Parametric Method

The instability of economic environment and the influence of exogenous uncertainties will inevitably impact the structure of economic variables.This change will result in the loss of the original test efficacy of the correlation test using the numerical characteristics of variables as the test means.At the same time,it will make the econometric model lose its original interpretation significance and application value.Therefore,the research purpose of this paper is to explore the effectiveness of root test of regression unit based on non-parametric method,which will help to identify the stability of economic variables under structural mutation.On the basis of introducing the basic theory of structural mutation,this paper presents the stability test of non-parametric regression.The non-parametric regression method can eliminate the structural trends in non-stationary variables and avoid the interference of structural mutations on the stability judgment of variables.The paper uses the R language to simulate the process of generating data containing different mutation types,and considers different mutation locations and mutation ranges.The stability of the identification data generation process is checked by using several types of stability.Firstly,it is proved that the existence of structural mutation can reduce the utility of traditional unit root test.Secondly,it is found that the error of structural mutation points can also result in the reduction of the effectiveness of the stability test of exogenous structures.Thirdly,the optimal window width is searched for the non-parameter degenerative stability test.Finally,the test efficacy of the stability test of non-parametric regression and the mutated unit of the endogenous structure are compared,and it is considered that the stability test of non-parametric regression has higher test efficacy.In terms of empirical research,the paper uses monthly inflation data from January 1990 to November 2018 for empirical analysis.Firstly,it is considered to be a non-stationary process using the traditional stability test.Secondly,it is found that the variable is a steady process containing structural mutation,which is very different from the traditional stability test.Therefore,it is necessary to consider structural mutations when studying the stability of variables.The innovation of this paper is mainly reflected in two aspects.Firstly,the results of this paper improve the efficiency of traditional stability test in identifying starvation of structural mutation variables.Secondly,the research provides a tool for identifying the stability of variables under the condition of structural mutation.This method can identify the stability of such variables more accurately.