Computer Dynamic Forecast Model with Adaptability through the Method of Rank-Sum Ratio

At present, global political and economic relations are very tense. After the occurrence of major political, economic and financial shocks, it will have a significant impact on the employment of all countries. Our goal is to analyze the impact of the trade war on China's imports and exports and on employment rate in the United States. Firstly, according to sino-us trade countermeasures and referring to SitcRev2, the goods were classified according to the added value of technology. Then, the coefficient of variation method was used to obtain the weight value. Secondly, the rank-sum ratio comprehensive evaluation method is used in to evaluate and rank the imports and exports of China and the United States from 2007 to 2018. Based on the above analysis and the factors affecting the employment rate, we establish a employment rate economic model. Through the collected data, VAR stationarity test and Granger causality test are conducted on the model. Finally, the specific impact of the trade war on the US employment rate is obtained through impulse response and variance decomposition.


Introduction
Trade disputes between China and the United States have recently become the focus of worldwide attention [1]. They concern not only China and the United States, but also the world economy and the world's political, military and diplomatic landscape. The issue of employment has always been the primary issue concerning the sound economic development and social stability of all countries in the world [2]. High-quality employment is an important part of a better life for the people and an inherent requirement for high-quality economic development. Therefore, it is crucial to build an economic model of employment rate.

Coefficient of variation method
According to SitcRev2, 238 commodities are classified according to the value added of technology [3]. In order to eliminate the influence of different dimensions of each evaluation index, it is necessary to use the coefficient of variation of each index to measure the difference degree of each index value [4]. The variation coefficient formula of each index is as follows: After calculation in SPSS, we get the weight value of China's import and export to the US, as shown in the figure below:

Rank-sum ratio comprehensive evaluation method
Based on the structural data of major commodity categories from 2007 to 2018, we adopted the ranksum ratio comprehensive evaluation method to evaluate the period from 2007 to 2018 [5]. The specific steps of the method are as follows: (1) Rank. The evaluation object of this paper is 2007-2018, and the evaluation index is 10 major commodity categories. The 10 evaluation indicators of 12 evaluation objects are arranged into a raw data table with 12 rows and 10 columns. The rank of each index evaluation object was compiled, in which the efficiency index was compiled from the smallest to the largest, the cost index was compiled from the largest to the small rank, and the average rank was compiled for those with the same index data. The resulting rank matrix is denoted as (2) Calculate the rank-sum ratio. According to the formula: When the weights of each evaluation index are different, the weighted rank sum ratio is calculated as follows: This process was carried out in MATLAB and the following results were obtained: Year rank of China's imports and exports to the United States.
According to Figure 2, we can intuitively see the lowest value in 2018, which indicates that the trade war has a great impact on China's imports and exports.

Employment rate economic model
Based on the above analysis and the factors affecting the employment rate, the following is the theoretical derivation of the relationship between each factor and the employment rate. First, suppose that the total output of a country conforms to the production function (1).
P is the total output, T is the technical factor, C is the capital, L is the labor force, M is the monopoly coefficient of the product, β and γ are the factor share coefficient, α is the efficiency that allows the factor to change the production process, and η is the monopoly ratio of the product. In order to maximize production, the product of marginal revenue of labor and wage ( w ) should be equal to the product of marginal revenue of capital and its use cost ( c ), that is, to eliminate the capital input in (1), we can get: When different countries produce different products, a country's domestic demand for commodities is met by both domestic and foreign production. When the equilibrium condition of supply equals demand is reached and the influence of exchange rate is considered, the following import and export equation can be obtained [6]:  (3) and (4), we can get: (2) and (5), we can get: By logarithmic processing on both sides of Equation (6), it can be obtained as follows: According to the employment rate calculation formula, it is assumed that the number of willing to work is 1 Y , the total number of population is Y , and the ratio of willing to work to the number of employed is x. Define the employment rate as follows [7]: The logarithm of both sides of equation (8) can be obtained as follows: Substitute equation (9) into equation (7)

VAR system stability test
First, in the EViews software we performed unit root checks on all variables and used information criteria to determine the lag order. Then the stability of VAR system is tested by eigenvalue method [8].
The results of running in EViews software are shown in Figure 3: All eigenvalues in the figure are located within the unit circle, so the VAR system is stable.

Granger causality test
Granger causality test method is adopted to test the causal relationship of each variable in the model, and part of the test results are shown in Figure 4.

Impulse response function analysis and variance analysis
In the EViews software, the dynamic impact effect of each factor on the employment rate of the United States is analyzed through the impulse response graph model, and the results were shown in the figure below [9]:  The variance decomposition method of prediction error is used to calculate the contribution ratio of each variable in the model to the mean square error forecast of the explained variable. The results of analysis of variance are as follows: