Bank potential credit client mining method based on knowledge graph and machine learning algorithm

Abstract

The invention provides a bank potential credit client mining method based on a knowledge graph and a machine learning algorithm. The method comprises the following steps: 1, a sample collection stage;2, a data preprocessing stage; 3, a model training stage. The method has the advantages that an efficient XGBoost integrated classifier is used for training a potential credit granting customer prediction model, potential credit granting customers with higher marketing success rate are mined, and accurate marketing of the customers is realized; the wide application and popularization of the method provide credit clients with higher marketing values for service personnel, the method improves the work efficiency of front-line service personnel, and has great significance and application valuesfor carrying out credit servicees in banks. A mode of extracting features based on the map is applied to potential credit granting customer mining for the first time, so that the application range ofthe knowledge map is expanded, and the development of the knowledge map is further promoted; and a good effect is achieved for solving the problem of nonuniform sample treatment.

Description

technical field [0001] The invention relates to a method for mining potential credit clients of a bank, in particular, an efficient, accurate, and uniform processing sample method for mining potential credit clients of a bank based on a knowledge map and a machine learning algorithm. Background technique [0002] Credit granting is a prerequisite for enterprises to apply for financing from banks. It is a prerequisite for using bank general loans, trade financing, and supply chain products. High-quality corporate credit granting customers have high stickiness and can bring interest and fee income to banks. . [0003] At present, the main ways for banks to develop credit customers are as follows: first, customers who have credit needs directly go to bank outlets and account managers to consult related services, which are called self-initiated customers; second, bank outlet credit account managers look for credit customers based on interpersonal relationships; There are dedica...

AI Technical Summary

Problems solved by technology

It is understood that although efficient and integrated machine learning algorithms are also widely used in banks, there are relatively few applications in mining potential credit customers, and there are few literatures based on the relationship feature mining of enterprise knowledge graphs.

Images