Interception strategy derivation method and system for credit anti-fraud

Abstract

The invention provides an interception strategy derivation method and system for credit and loan anti-fraud, which can quickly derive an effective interception strategy from massive data and identify users with high overdue risks, and comprises the following steps: acquiring sample data, extracting characteristic variables related to the users from the sample data, and performing data preprocessing, performing binning processing on the characteristic variables to obtain variable binning; calculating a WOE value and an IV value of the characteristic variable under each variable sub-box; wOE coding is carried out, and a WOE value is used to replace the numerical value of the characteristic variable corresponding to the variable binning; a prediction model is established and trained, the prediction model is evaluated through an evaluation index AUC, model parameters are adjusted, and an optimal prediction model is obtained; calculating scores of all variable sub-boxes through the optimal prediction model, and crossing the variable sub-boxes of which the scores are not greater than a set threshold value to generate an interception strategy; and verifying whether the interception strategies meet online conditions or not, and reserving all the interception strategies meeting the online conditions.

Description

technical field [0001] The present invention relates to the technical field of big data credit risk control, and in particular to a derivation method, system, computer device and computer-readable storage medium for interception strategy for credit anti-fraud. Background technique [0002] With the rapid development of the Internet, more and more industries and companies are actively integrating into the new century Internet model. Among them, Internet finance is slowly entering door-to-door, and credit lending based on personal and corporate big data information is quietly changing traditional bank lending. However, with the development of technology and the simplification of Internet lending methods, more and more new fraudulent loan fraud methods have appeared, causing immeasurable losses to the current Internet and bank credit business. [0003] The rise of big data has made it easier to obtain information, and it has given a direction to fight against fraud and loan ga...

AI Technical Summary

Problems solved by technology

However, in the face of the complex and changeable business environment and the terabyte or even petabyte data scale in the era of big data, the traditiona

Images