Method for judging relevance between quota and credit granting amount based on statistical test of concatenated table

Abstract

The invention discloses a method for judging the relevance between a quota and a credit granting amount based on a column connection table statistical test, namely judging the relevance between the quota and the credit granting amount through a chi-square test method, and quantitatively counting the deviation degree between an actual observation value and a theoretical value of a sample. When statistical test is carried out, the chi-square value is compared with the corresponding critical value, and a related statistical conclusion can be obtained. The method for judging the relevance between the quota and the credit granting amount based on the statistical test of the column connection table comprises a chi-square test method and a Fisher accurate test method. The solution is based on solid mathematical and statistical theories, the defect that the judgment of graphs and correlation coefficients is weakened under the condition of big data is overcome, and the quantitative fuzziness of the graph trend is avoided; the scientific judgment on the relevance between the variables is facilitated; the method is suitable for a numerical type and can also be used for classifying variables.

Description

technical field [0001] The invention relates to the technical field of algorithms, in particular to a method for judging the correlation between the amount raised and the credit amount based on the statistical test of the contingency table. Background technique [0002] Lending institutions usually hope that users can increase the credit amount in the case of granting credit to users, so as to increase the institution's lending scale and expand profits. [0003] In a general sense, practitioners tend to think that there is a certain degree of correlation between the credit limit and the credit limit, that is, users with higher credit limits are more inclined to increase the limit, and vice versa, users with low credit limits are more likely to increase the limit. Small. However, the main analysis method at present is still based on the judgment of the graphic trend, supplemented by the way of naked eye recognition. [0004] From the perspective of data analysis, when data ...

AI Technical Summary

Problems solved by technology

[0005] Although the drawing and correlation coefficient are simple to understand and highly interpretable, there are still some bottlenecks

For example, graphs such as trend charts and comparison charts will be affected by the coordinate range, that is, different coordinate ranges will lead to differences in visual perception, resulting in deviations in decision-making and judgment

For another example, if one of the two variables is a categorical variable, the correlation coefficient cannot reflect the relationship between the variables

In addition, although the graphical trend can make qualitative judgments, it is insufficient for situations that require quantitative judgments

Images