Credit evaluation method and device based on big data, electronic device and storage medium

Abstract

The embodiment of the invention provides a credit evaluation method and device based on big data, an electronic device and a storage medium, which relate to the technical field of big data. The methodcomprises the following steps of dividing a plurality of index variables of a sample user into a plurality of equal frequency intervals by using an equal frequency method; taking an end point of theequal frequency interval as an initial clustering center to perform clustering operation on the plurality of index variables to obtain a plurality of variable boxes of each index variable; extractingvariable features from the variable sub-boxes of each index variable, and training the credit evaluation model based on the variable features; evaluating the credit of the user to be evaluated by thetrained credit evaluation model. The technical proposal of the embodiment of the invention can more accurately reflect the variable value distribution characteristics of different index variables, thereby improving the accuracy of credit evaluation model evaluation.

Description

technical field [0001] The present invention relates to the technical field of big data, and in particular, relates to a credit evaluation method, a credit evaluation device, electronic equipment, and a computer-readable storage medium. Background technique [0002] With the development of Internet finance, more and more people purchase financial products through online platforms. In order to reduce business risks, financial institutions need to evaluate the credit of users. [0003] In the existing credit scoring model, the variables in the sample data need to be binned or discretized. The main binning methods include equidistant binning and equal frequency binning. Equidistant binning means that each bin after division The interval values ​​of the bins are equal, and equal-frequency binning means that the amount of data in each bin is equal after division. However, these two binning methods are difficult to reflect the variable value distribution characteristics of differe...

AI Technical Summary

Problems solved by technology

[0003] In the existing credit scoring model, the variables in the sample data need to be binned or discretized. The main binning methods include equidistant binning and equal frequency binning. Equidistant binning means that each bin after division The interval values ​​of the bins are equal, and equal-frequency binning means that the amount of data in each bin is equal after division. However, these two binning methods are difficult to reflect the variable value distribution characteristics of different variables, making the evaluation results of the information evaluation model accurate. low rate

Images