Abstract:In order to help banks formulate Credit Strategies for small, medium-sized and micro enterprises, this paper analyzes enterprises from two aspects. On the one hand, starting from the enterprise's data and from the enterprise's relevant data, including qualitative indicators such as reputation level, invoice information, credit record and operating income, a model is established to quantitatively analyze the operation status and reputation status of the enterprise. On the other hand, through the analysis of the "maximum lending rate" and "maximum lending rate" of the bank, we can analyze the "maximum lending rate" and "how much lending rate" of the bank. Combining the above variables, we can get the best credit strategy of the bank under the factors of credit risk and income. Firstly, without considering the influence of other data, this paper abstracts five indicators to measure the reputation status and enterprise-scale from the data of small and medium-sized enterprises with credit rating, quantifies the indicators. At the same time, the Euclidean distance is calculated by MATLAB modeling, the correlation between lending enterprises and non-lending enterprises is analyzed, and the high-risk enterprises in lending enterprises are selected. Then, among the remaining enterprises that can issue loans, five different credit strategies are formulated according to the weighting of their credit rating and enterprise scale as the primary basis for the number of bank loans and interest rate. Secondly, the focus is on how to comprehensively quantify the credit status and business problems of small, medium-sized and micro enterprises without credit records. Without considering the influence of other data, this paper mainly analyzes the correlation between the enterprises without credit records and the enterprises mentioned above that do not grant loans. It preliminarily obtains the enterprises that can grant loans. Then, through decision tree modeling, the loan demand of enterprises with different credit levels is estimated as a variable. By fitting the curve between interest rate and customer churn rate, different interest rate ranges under different levels are obtained as the second variable. Finally, through the above two variables, we establish the profit function of bank credit, find the corresponding variables at the maximum value, and formulate relevant credit strategies.

Credit risk assessment of small and micro enterprise based on machine learning

Research and Design of Credit Risk Assessment System Based on Big Data and Machine Learning

Using a genetic backpropagation neural network model for credit risk assessment in the micro, small and medium-sized enterprises

A two-stage credit scoring model based on random forest: Evidence from Chinese small firms

Research on credit strategy of small, medium, and micro enterprises based on commercial banks

Establishing a Credit Risk Evaluation System for SMEs Using the Soft Voting Fusion Model

Predicting the credit risk of small and medium‐sized enterprises in supply chain finance using machine learning algorithms

Comparative study of credit rating models for small and micro enterprises

Multimodel Integrated Enterprise Credit Evaluation Method Based on Attention Mechanism

Advanced User Credit Risk Prediction Model using LightGBM, XGBoost and Tabnet with SMOTEENN

Machine Learning-Based Risk Management of Credit Sales in Small and Midsize Business

Credit Decision Problems of MSMEs (Medium, Small and Micro-sized Enterprises)

Credit Strategy Design of Small and Medium-Sized Enterprises

The Evaluation Model of Enterprise Credit Default Based on Logistic Regression

Credit risk assessment model of small and medium-sized enterprise based on logistic regression

Enterprise Financial Risk Analysis Based on Improved Model C-Means Clustering Algorithm

Impact of Enterprises Relationship on Ml's Ability to Forecast SME's Credit Risk in SCF: A Research Based on Complex Network

Feature Enhanced Ensemble Modeling with Voting Optimization for Credit Risk Assessment

CatBoost model with synthetic features in application to loan risk assessment of small businesses

Credit Risk Simulation of Enterprise Financial Management Based on Machine Learning Algorithm

A Multicriteria Approach for Modeling Small Enterprise Credit Rating: Evidence from China