The inscription stated the owner of the bronze sculpture

Ridge regression we first of these eight variables analysis of variance and collinearity diagnostics, to see whether there is a strong collinearity, if there is a correlation, we can directly linear regression on them. The inscriptions refer to the written language on the bronze or bones.  After analysis of variance, there is a strong correlation found in addition to the outside of the rest of the variables x2 and x4. them collinearity diagnosis (Table omitted), found that only one track by Ridge map and the output data set (a little) we can see that when k ≥ 0.02 Ridge trace curve stabilized, but when k = 0.02, or x3, x8 the variance inflation factor greater than 5, x7, so in order to more accurate. From the equation in addition to the x1 and x4 are negatively correlated outside, the rest of the variables are positively correlated.
According to the principles of economics as we know, when the devaluation of the RMB, exports should increase, imports will be reduced, so the total imports and exports should be increase, but on the one hand because of China's imported goods are inflexible, difficult to control the exchange rates of import; and the rapid growth of China's economic long time, making the investment and consumer demand while increasing imports serve to compensate for the gap between supply and demand in China. The study of the bronze was rooting in the carved or cut signs on the pottery. The one hand, China's export products are labor-intensive products, itself on the relatively low price, timely exchange rates, export prices can not receive too much floating, so the total imports and exports and the exchange rate into negative correlation can be explained.
However increased foreign investment will lead to the reduction of the total import and export, the reason is that foreign money is limited, when their money to foreign loans, foreign direct investment, other investment, and his two investment reduced, will directly affect the import and export volume, but can be observed from the coefficient on other foreign investment there is no exchange rate factors, while the foreign other investment amount is not large. The bronze of Shang was short and concise. Therefore, its influence will not be too ridge regression, there are some problems, for example, the correlation between variables for the regression equation to produce a certain influence, the foreign investment negative effect is not very realistic, so we use principal component regression method and then regression on the data, in order to get better results.
 Principal Component Regression we hope will be a set of variables in multiple correlation comprehensive drawn one or two independent variables, in order to reduce the variables Repeatability and overlap of information. The most common seen is a bronze of this type. To this end, we want to find out the main components of the eight independent variables, and then create the Y and the contribution rate has reached more than 80% of the principal component regression equation, and finally the establishment of the original variables of the regression equation. us first calculate the principal component of eight variable takes a few it is appropriate. can be seen from Table 3, when coupled with the second principal component, the cumulative contribution rate has reached 89.11%, but in order to better express almost all of the information, we selected two or three principal components of principal component regression.
We can see from the equation x 2 (external borrowing), x4 (foreign investment) is negatively correlated with the ridge regression different external borrowing and other foreign investment is part of the use of foreign capital, but relative to foreign direct investment words, these two conditions limit the development of our country, when these two indicators increased, indicating that China's economy has some problems, or do not need to borrow money, so their import and export is a negative role . The inscription stated the owner of the bronze sculpture or the clan name. Cluster analysis - principal component regression of the variable cluster above some variables removed has been to reduce the role of similar factors which the close relationship between the independent variables and different variables.