Overdue loan customer grouping method

Abstract

The invention relates to the technical field of loan overdue customer grouping, and discloses a loan overdue customer grouping method. A customer grouping importance set is formed by acquiring repayment behaviors, basic information, credit scores, loan amounts and purposes of users, a score formula and an initial weight coefficient of each element are designed, a weight adjustment strategy is combined, and finally the importance and the weight of each element are fused, so that the user comprehensive scoring method is obtained. Through multi-dimensional information integration, a financial institution can scientifically carry out user grouping and risk control, and the accuracy of loan approval and the quality of personalized financial services are improved. A weight distribution and adjustment strategy ensures scientificity and rationality of evaluation, and excessive influence of a single factor on an evaluation result is avoided, so that comprehensive evaluation and accurate classification management of the user credit status are realized. Therefore, all aspects of the credit status of the user can be comprehensively and systematically covered, and the comprehensiveness and accuracy of evaluation can be ensured.

Description

Technical Field

[0001] The present invention relates to the technical field of overdue loan customer grouping, and in particular to a method for grouping overdue loan customers. Background Art

[0002] With the development of financial markets and the change of consumption concepts, borrowing has become a widespread economic activity in modern society. However, the problem of overdue loans has become increasingly prominent, seriously affecting the asset quality and market stability of financial institutions. Overdue loans not only increase the risks of financial institutions, but may also lead to instability in the entire financial system. Therefore, how to effectively assess and manage the risks of borrowers and reduce the occurrence of overdue loans is an important issue facing financial institutions.

[0003] Traditional risk assessment methods for borrowers mainly rely on single-dimensional information, such as credit scores or income levels. Although these methods can reflect the repayment ability and willingness of customers to a certain extent, they often ignore the comprehensive impact of multi-dimensional factors on customer risk. For example, factors such as the borrower's career stability, loan purpose, and repayment behavior also have an important impact on the customer's repayment ability and overdue risk. The singleness and limitations of traditional assessment methods make it impossible to comprehensively and accurately assess the risk level of customers in practical applications. This case intends to propose a method for comprehensive risk assessment and grouping of customers based on multi-dimensional factors. Summary of the invention

[0004] The present invention provides a method for grouping customers with overdue loans, which helps solve the problems mentioned in the above background technology.

[0005] The present invention provides the following technical solution: a method for grouping overdue loan customers, comprising:

[0006] Get the user's repayment behavior, denoted as I B , the repayment behavior includes the number of on-time repayments, total repayments and overdue payments in the last 12 months;

[0007] Get basic user information, denoted as I D , the user's basic information includes annual income, age and occupation;

[0008] Get the user's credit score, denoted as I C ;

[0009] Get the user's loan amount, denoted as I A ;

[0010] Get the user's loan purpose, denoted as I U ;

[0011] Will I B ,I D ,IC ,I A ,I U Composing customer segment importance sets;

[0012] Design a scoring formula for each element in the customer segmentation importance set;

[0013] Design the initial weight coefficient of each element based on the importance of each element in the customer group importance set;

[0014] Design weight adjustment strategies;

[0015] The importance and weight of each element in the customer group importance set are integrated to obtain a comprehensive user scoring method;

[0016] Use customer segmentation strategies to segment customers.

[0017] Optionally, the scoring formula for each element in the customer group importance set is designed to specifically include:

[0018] Repayment Behavior I B =2×R pay +3×(1-R default );

[0019] The R pay R is the ratio of the number of on-time repayments to the total number of repayments in the last 12 months. default The ratio of the expected number of repayments to the total number of repayments in the last 12 months;

[0020] Basic InformationI D =0.5×I income +0.3×A age +0.2×P profession ;

[0021] I income A is the user's annual income divided by the benchmark income. age P is the user's age divided by the base age. profession Rate careers on career stability and earning potential;

[0022] User Credit Score I C =S credit ;

[0023] The S credit For credit scores, credit scores are provided directly by credit scoring agencies, ranging from 300 to 850 divided by 850;

[0024] User loan amount I A =0.7×A loan +0.3×(1-R income );

[0025] The A loan The loan amount is divided by the base amount, R income The ratio of the loan amount to annual income;

[0026] User loan purpose I U =0.6×(1-R default )+0.4×U purpose ;

[0027] The R default is the ratio of overdue payments to total repayments in the last 12 months, U purpose Score the purpose of the loan.

[0028] Optionally, the career stability and income potential score may include:

[0029] According to occupational stability, occupations are divided into high-stability occupations, medium-stability occupations, and low-stability occupations.

[0030] The elements in the high-stability occupation set are civil servants, teachers, medical professionals and business managers;

[0031] The elements in the medium stability occupation set are financial analysts, engineers and technical support;

[0032] The elements in the low-stability occupation set are freelancers, sales representatives, contract workers and temporary workers;

[0033] According to income potential, occupations are divided into a high-income potential occupation set, a medium-income potential occupation set and a low-income potential occupation set;

[0034] The elements of the high-income potential occupation set are senior managers, doctors, financial industry executives, technology industry executives and top engineers;

[0035] The elements of the medium-income potential occupation set are middle-level managers, lawyers, and senior technicians;

[0036] The elements of the low-income potential occupation set are administrative assistants, service industry workers, and sales assistants;

[0037] Let high stability occupations have a score of 3, medium stability occupations have a score of 2, and low stability occupations have a score of 1;

[0038] Assume that the score of high-income potential occupations is 3, the score of medium-income potential occupations is 2, and the score of low-income potential occupations is 1;

[0039] Let the stability weight be A, the income weight be B, and A+B=1;

[0040] Then the career score = (stability score × A) + (income potential score × B).

[0041] Optionally, the loan purpose score specifically includes:

[0042] Set a loan purpose set, the elements in the set are consumer loan C1, housing loan C2, car loan C3, education loan C4, business loan C5, agricultural loan C6, investment loan C7 and emergency loan C8;

[0043] The scoring formula for loan purpose is Score(C i )=w1×Risk(C i )+w2×Amount(C i )+w3×Repayment(Ci), where C i represents the i-th element in the loan purpose set, Risk(C i ) indicates category C i The risk level, Amount(C i ) represents the loan amount, Repayment(Ci) represents the repayment ability, w1, w2, w3 represent the weights of risk, amount, and repayment ability respectively, and w1+w2+w3=1.

[0044] Optionally, the initial weight coefficients of the design elements specifically include:

[0045] The importance index of each element is standardized to between 0 and 1, then Among them I i is the importance index of the i-th element, is the importance after normalization;

[0046] Initial weight coefficient α i It can be calculated based on the standardized importance index, This ensures that all initial weight coefficients sum to 1.

[0047] Optionally, the design weight adjustment strategy specifically includes:

[0048] The design weight adjustment formula is: where ω i is the weight of the i-th element in the importance set, α i is the initial weight coefficient of the i-th element, I i is the importance score of the ith element, is the weighted sum of all elements and is used to normalize the weights so that the sum of all weights is 1.

[0049] Optionally, the user comprehensive scoring method specifically includes:

[0050] The weighted sum of each element in the importance set is: S = ∑ i ω i ×Score i where ω i The weight of the i-th element, Score i Get the importance score for the i-th element.

[0051] Optionally, the customer grouping strategy specifically includes:

[0052] Divide user risk levels according to their comprehensive scores;

[0053] Very low risk: S ≥ 80;

[0054] Low risk: 70≦S<80;

[0055] Medium risk: 60≦S<70;

[0056] High risk: 50≦S<60;

[0057] Very high risk: S<50;

[0058] Set different risk level sets;

[0059] First, match the risk level according to the user's specific score, and then add the user to the corresponding risk level set.

[0060] The present invention has the following beneficial effects:

[0061] 1. By obtaining the user's repayment behavior information, especially the repayment data within the last 12 months, the user's credit reliability and repayment habits can be effectively evaluated; the records of the number of on-time repayments and the total number of repayments can help financial institutions understand the user's repayment ability and willingness, and the number of overdue payments directly reflects the user's default risk; obtaining the user's basic information such as annual income, age and occupation can provide important background information for credit assessment; annual income is a direct indicator of the user's repayment ability, age can reflect the user's stability and life cycle stage, and occupation can further subdivide the user's income stability and occupational risk; obtaining the user's credit score is a brief summary of his or her overall credit status; credit scores are usually calculated based on the user's credit history, repayment behavior and other relevant factors, and can intuitively reflect the user's credit risk level; obtaining the user's loan amount can help financial institutions understand the user's borrowing needs and the scale of funds used; the loan amount is one of the important factors that determine the user's repayment pressure. By combining the user's income, credit score and repayment behavior, it can be more Accurately assess the user's repayment ability and risk level; by understanding the purpose of the loan, the risk of the loan can be more specifically assessed, and corresponding management and control measures can be formulated; this can not only improve the accuracy of loan approval, but also help financial institutions better understand user needs, thereby providing more personalized financial services; forming a customer segmentation importance set, it can comprehensively and systematically segment and assess the risk of users; this set method makes the assessment process more structured and systematic, and can cover all aspects of the user's credit status to ensure the comprehensiveness and accuracy of the assessment; through this multi-dimensional information integration, financial institutions can more scientifically segment and control users; designing a scoring formula for each element in the customer segmentation importance set can convert complex user information into a quantitative scoring indicator; designing the initial weight coefficient based on the importance of each element in the customer segmentation importance set can ensure that each factor in the assessment process is properly valued; the setting of the weight coefficient can reflect the relative importance of each element in the overall assessment, ensuring the scientificity and rationality of the assessment results. Through reasonable weight allocation, the user's comprehensive credit status can be reflected more accurately, avoiding the excessive influence of a single factor on the evaluation results; through the weight adjustment strategy, the weight of each element can be adjusted in time according to the actual data and changes in the market environment to ensure the accuracy and effectiveness of the evaluation results; by integrating the importance and weight of each element in the customer grouping importance set to obtain a comprehensive user scoring method, a comprehensive and systematic evaluation of the user's credit status can be achieved; using customer grouping strategies to group customers can achieve accurate classification and management of users.

[0062] 2. R pay It reflects the proportion of users who made on-time repayments in the past year and is an important indicator for measuring users’ credit willingness. defaultIt is a direct reflection of overdue behavior and can reveal the credit risk of users. By clarifying the calculation methods of these two ratios, the repayment behavior assessment can be made more operational and transparent. income By comparing the user's annual income with the benchmark income, it reflects the user's relative economic ability; age By comparing the user's age with the benchmark age, the user's life cycle stage is reflected; profession Scoring by career stability and income potential reveals the reliability of the user's career and future income expectations; credit The credit scores provided by credit scoring agencies, ranging from 300 to 850, are divided by 850 to ensure the standardization and reliability of the scores. By adopting the scoring data of authoritative agencies, the efficiency and accuracy of the evaluation process can be greatly improved. loan By comparing the loan amount with the benchmark amount, the relative loan size of the user is reflected; R income The user's repayment ability is evaluated by the ratio of the loan amount to the annual income; the standardized calculation method can ensure the comparability of the loan amount evaluation between different users, and improve the fairness and reliability of the evaluation results; default The proportion of overdue payments to total repayments reflects the user's default risk; purpose The actual use and rationality of the loan can be evaluated by scoring the loan purpose. The standardized calculation method can ensure the comparability of loan purpose evaluations between different users, and improve the fairness and reliability of the evaluation results.

[0063] 3. Classifying occupations into high-stability occupations, medium-stability occupations and low-stability occupations according to occupational stability helps to classify and evaluate users' credit risks from an occupational perspective; occupational stability is an important indicator of users' income and employment security; high-stability occupations usually mean more stable income and lower unemployment risk, while low-stability occupations may face greater income fluctuations and employment uncertainty; through such classification, users' repayment ability and default risk can be more accurately assessed, thereby providing more targeted analysis and decision-making basis in the credit assessment process; according to income potential, occupations are divided into high-income potential occupations, medium-income potential occupations and low-income potential occupations. The occupational set can be used to classify and evaluate the user's credit risk from the perspective of income prospects; occupations with high income potential usually mean higher expectations for future income growth, while occupations with low income potential may face bottlenecks in income growth; through such classification, the user's long-term repayment ability and financial stability can be more accurately evaluated; setting corresponding scores for occupations with different stability and different income potential can quantify the impact of occupational stability on credit assessment; setting stability weight A and income weight B, and ensuring A+B=1, can balance the impact of occupational stability and income potential on credit assessment; such weight settings can be adjusted according to actual needs to ensure the flexibility and adaptability of the assessment model.

[0064] 4. Set a set of borrowing purposes and list in detail eight borrowing purposes, including consumer loans, housing loans, car loans, education loans, operating loans, agricultural loans, investment loans and emergency loans, which will help to classify and analyze different types of loans. Different loan purposes often have different risk characteristics and repayment capacity requirements. Through clear classification, the risk and repayment capacity of each loan purpose can be evaluated more specifically, thereby improving the scientificity and accuracy of loan approval. By comprehensively considering the risk level, loan amount and repayment capacity, quantitative evaluation is conducted on different borrowing purposes. By setting weights w1, w2, w3 and ensuring that their sum is 1, the influence weights of different factors can be dynamically adjusted according to actual conditions, thereby more flexibly adapting to market changes and risk management needs. This will not only improve the accuracy and adaptability of the scoring model, but also help financial institutions make more scientific decisions in the loan approval process.

[0065] 5. Standardizing the importance index of each element to between 0 and 1 allows different indicators to be compared and calculated on the same scale, thereby eliminating the impact of dimensional differences in each indicator; through standardization, it is possible to avoid certain indicators from dominating the calculation due to their large original value range, thereby achieving fair comparison of each indicator; the initial weight coefficient is calculated based on the standardized importance index, which can ensure that the sum of all weight coefficients is 1, thereby ensuring that the total weight is constant; by calculating the initial weight based on the standardized importance index, it can ensure that the actual importance of each indicator is fully reflected in the total weight, thereby improving the scientificity and rationality of the model.

[0066] 6. By designing a weight adjustment formula, the weights can be dynamically adjusted according to the importance scores of each dimension, making the model more flexible and adaptable. The use of a formulaic calculation method can reduce the impact of human factors on weight setting. The formula uses an exponential function and normalization processing to standardize the weights to a value between 0 and 1, ensuring that the sum of all weights is 1. This standardization process makes the weight setting more standardized. By introducing importance scores, the formula can fully consider the differences in importance of different dimensions in risk assessment.

[0067] 7. The weighted sum of each element in the importance set is calculated to obtain a comprehensive score S, which provides a comprehensive and quantitative result for the evaluation model. In this way, the influence of each importance element can be comprehensively considered to obtain a unified scoring result. The weighted sum can fully reflect the role of each element in the overall evaluation and ensure the comprehensiveness and scientificity of the evaluation results. i The weight of the i-th element determines the influence of the element in the comprehensive score; reasonable weight distribution can ensure that the contribution of each element in the total score matches its importance, avoiding some elements from being overly prominent or neglected.

[0068] 8. Classifying users into risk levels based on their comprehensive scores is a key step in converting quantitative scoring results into actionable risk classifications. This method can help financial institutions quickly and accurately identify and classify users' credit risk status, thereby better managing loan approval and risk control processes. By clearly classifying risk levels, financial institutions can formulate corresponding strategies and measures for users of different risk levels to improve the effectiveness and accuracy of risk management. Setting different risk level sets and classifying various types of users into different categories such as extremely low risk, low risk, medium risk, high risk and extremely high risk can help financial institutions to systematically manage and classify users. First, matching the corresponding risk level according to the user's specific score, and then adding the user to the corresponding risk level set can ensure the accuracy and standardization of the risk assessment process. BRIEF DESCRIPTION OF THE DRAWINGS

[0069] Figure 1 It is a schematic diagram of the process of the present invention. DETAILED DESCRIPTION

[0070] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0071] Example, see Figure 1 , a method for grouping overdue loan customers, comprising:

[0072] Get the user's repayment behavior, denoted as I B , the repayment behavior includes the number of on-time repayments, total repayments and overdue payments in the last 12 months;

[0073] Get basic user information, denoted as I D , the user's basic information includes annual income, age and occupation;

[0074] Get the user's credit score, denoted as I C ;

[0075] Get the user's loan amount, denoted as I A ;

[0076] Get the user's loan purpose, denoted as I U ;

[0077] Will I B ,I D ,I C ,I A ,I U Composing customer segment importance sets;

[0078] Design a scoring formula for each element in the customer segmentation importance set;

[0079] Design the initial weight coefficient of each element based on the importance of each element in the customer group importance set;

[0080] Design weight adjustment strategies;

[0081] The importance and weight of each element in the customer group importance set are integrated to obtain a comprehensive user scoring method;

[0082] Use customer segmentation strategies to segment customers.

[0083] By obtaining the user's repayment behavior information, especially the repayment data within the last 12 months, the user's credit reliability and repayment habits can be effectively evaluated; the records of the number of on-time repayments and the total number of repayments can help financial institutions understand the user's repayment ability and willingness, and the number of overdue payments directly reflects the user's default risk; obtaining the user's basic information such as annual income, age and occupation can provide important background information for credit assessment; annual income is a direct indicator of the user's repayment ability, age can reflect the user's stability and life cycle stage, and occupation can further subdivide the user's income stability and occupational risk; obtaining the user's credit score is a brief summary of his or her overall credit status; credit scores are usually calculated based on the user's credit history, repayment behavior and other relevant factors, and can intuitively reflect the user's credit risk level; obtaining the user's loan amount can help financial institutions understand the user's borrowing needs and the scale of funds used; the loan amount is one of the important factors that determine the user's repayment pressure. By combining the user's income, credit score and repayment behavior, it can be more accurate. Accurately assess the user's repayment ability and risk level; by understanding the purpose of the loan, the risk of the loan can be more specifically assessed, and corresponding management and control measures can be formulated; this can not only improve the accuracy of loan approval, but also help financial institutions better understand user needs, thereby providing more personalized financial services; forming a customer segmentation importance set, it can comprehensively and systematically segment and assess the risk of users; this set method makes the assessment process more structured and systematic, and can cover all aspects of the user's credit status to ensure the comprehensiveness and accuracy of the assessment; through this multi-dimensional information integration, financial institutions can more scientifically segment and control users; designing a scoring formula for each element in the customer segmentation importance set can convert complex user information into a quantitative scoring indicator; designing the initial weight coefficient based on the importance of each element in the customer segmentation importance set can ensure that each factor in the assessment process is given proper attention; the setting of the weight coefficient can reflect the relative importance of each element in the overall assessment, ensuring the scientificity and rationality of the assessment results. Through reasonable weight allocation, the user's comprehensive credit status can be reflected more accurately, avoiding the excessive influence of a single factor on the evaluation results; through the weight adjustment strategy, the weight of each element can be adjusted in time according to the actual data and changes in the market environment to ensure the accuracy and effectiveness of the evaluation results; by integrating the importance and weight of each element in the customer grouping importance set to obtain a comprehensive user scoring method, a comprehensive and systematic evaluation of the user's credit status can be achieved; using customer grouping strategies to group customers can achieve accurate classification and management of users.

[0084] The scoring formula for each element in the design customer group importance set specifically includes:

[0085] Repayment Behavior IB =2×R pay +3×(1-R default );

[0086] The R pay R is the ratio of the number of on-time repayments to the total number of repayments in the last 12 months. default The ratio of the expected number of repayments to the total number of repayments in the last 12 months;

[0087] Basic InformationI D =0.5×I income +0.3×A age +0.2×P profession ;

[0088] I income A is the user's annual income divided by the benchmark income. age P is the user's age divided by the base age. profession Rate careers on career stability and earning potential;

[0089] User Credit Score I C =S credit ;

[0090] The S credit For credit scores, credit scores are provided directly by credit scoring agencies, ranging from 300 to 850 divided by 850;

[0091] User loan amount I A =0.7×A loan +0.3×(1-R income );

[0092] The A loan The loan amount is divided by the base amount, R income The ratio of the loan amount to annual income;

[0093] User loan purpose I U =0.6×(1-R default )+0.4×U purpose ;

[0094] The R default is the ratio of overdue payments to total repayments in the last 12 months, U purpose Score the purpose of the loan.

[0095] R pay It reflects the proportion of users who made on-time repayments in the past year and is an important indicator for measuring users’ credit willingness. defaultIt is a direct reflection of overdue behavior and can reveal the credit risk of users. By clarifying the calculation methods of these two ratios, the repayment behavior assessment can be made more operational and transparent. income By comparing the user's annual income with the benchmark income, it reflects the user's relative economic ability; age By comparing the user's age with the benchmark age, the user's life cycle stage is reflected; profession Scoring by career stability and income potential reveals the reliability of the user's career and future income expectations; credit The credit scores provided by credit scoring agencies, ranging from 300 to 850, are divided by 850 to ensure the standardization and reliability of the scores. By adopting the scoring data of authoritative agencies, the efficiency and accuracy of the evaluation process can be greatly improved. loan By comparing the loan amount with the benchmark amount, the relative loan size of the user is reflected; R income The user's repayment ability is evaluated by the ratio of the loan amount to the annual income; the standardized calculation method can ensure the comparability of the loan amount evaluation between different users, and improve the fairness and reliability of the evaluation results; default The proportion of overdue payments to total repayments reflects the user's default risk; purpose The actual use and rationality of the loan can be evaluated by scoring the loan purpose. The standardized calculation method can ensure the comparability of loan purpose evaluations between different users, and improve the fairness and reliability of the evaluation results.

[0096] The career stability and income potential scores include:

[0097] According to occupational stability, occupations are divided into high-stability occupations, medium-stability occupations, and low-stability occupations.

[0098] The elements in the high-stability occupation set are civil servants, teachers, medical professionals and business managers;

[0099] The elements in the medium stability occupation set are financial analysts, engineers and technical support;

[0100] The elements in the low-stability occupation set are freelancers, sales representatives, contract workers and temporary workers;

[0101] According to income potential, occupations are divided into a high-income potential occupation set, a medium-income potential occupation set and a low-income potential occupation set;

[0102] The elements of the high-income potential occupation set are senior managers, doctors, financial industry executives, technology industry executives and top engineers;

[0103] The elements of the medium-income potential occupation set are middle-level managers, lawyers, and senior technicians;

[0104] The elements of the low-income potential occupation set are administrative assistants, service industry workers, and sales assistants;

[0105] Let high stability occupations have a score of 3, medium stability occupations have a score of 2, and low stability occupations have a score of 1;

[0106] Assume that the score of high-income potential occupations is 3, the score of medium-income potential occupations is 2, and the score of low-income potential occupations is 1;

[0107] Let the stability weight be A, the income weight be B, and A+B=1;

[0108] Then the career score = (stability score × A) + (income potential score × B).

[0109] Classifying occupations into high-stability occupations, medium-stability occupations and low-stability occupations according to occupational stability helps to classify and evaluate users' credit risks from an occupational perspective; occupational stability is an important indicator of users' income and employment security; high-stability occupations usually mean more stable income and lower unemployment risks, while low-stability occupations may face greater income fluctuations and employment uncertainties; through such classification, users' repayment ability and default risks can be assessed more accurately, thereby providing more targeted analysis and decision-making basis in the credit assessment process; according to income potential, occupations are divided into high-income potential occupations, medium-income potential occupations and low-income potential occupations. By combining different occupations, the credit risk of users can be classified and assessed from the perspective of income prospects. Occupations with high income potential usually mean higher expectations for future income growth, while occupations with low income potential may face bottlenecks in income growth. Through such classification, users' long-term repayment ability and financial stability can be assessed more accurately. By setting corresponding scores for occupations with different stability and different income potential, the impact of occupational stability on credit assessment can be quantified. By setting stability weight A and income weight B and ensuring that A+B=1, the impact of occupational stability and income potential on credit assessment can be balanced. Such weight settings can be adjusted according to actual needs to ensure the flexibility and adaptability of the assessment model.

[0110] The loan purpose score specifically includes:

[0111] Set a loan purpose set, the elements in the set are consumer loan C1, housing loan C2, car loan C3, education loan C4, business loan C5, agricultural loan C6, investment loan C7 and emergency loan C8;

[0112] The scoring formula for loan purpose is Score(C i)=w1×Risk(C i )+w2×Amount(C i )+w3×Repayment(Ci), where C i represents the i-th element in the loan purpose set, Risk(C i ) indicates category C i The risk level, Amount(C i ) represents the loan amount, Repayment(Ci) represents the repayment ability, w1, w2, w3 represent the weights of risk, amount, and repayment ability respectively, and w1+w2+w3=1.

[0113] Setting a set of borrowing purposes and listing in detail eight borrowing purposes, including consumer loans, housing loans, car loans, education loans, operating loans, agricultural loans, investment loans and emergency loans, will help to classify and analyze different types of loans; different loan purposes often have different risk characteristics and repayment capacity requirements. Through clear classification, the risk and repayment capacity of each loan purpose can be evaluated more specifically, thereby improving the scientificity and accuracy of loan approval; by comprehensively considering the risk level, loan amount and repayment capacity, different borrowing purposes are quantitatively evaluated; by setting weights w1, w2, w3 and ensuring that their sum is 1, the influence weights of different factors can be dynamically adjusted according to actual conditions, thereby more flexibly adapting to market changes and risk management needs; this will not only improve the accuracy and adaptability of the scoring model, but also help financial institutions make more scientific decisions in the loan approval process.

[0114] The initial weight coefficients of the design elements specifically include:

[0115] The importance index of each element is standardized to between 0 and 1, then Among them I i is the importance index of the i-th element, is the importance after normalization;

[0116] Initial weight coefficient α i It can be calculated based on the standardized importance index, This ensures that all initial weight coefficients sum to 1.

[0117] Standardizing the importance index of each element to between 0 and 1 allows different indicators to be compared and calculated on the same scale, thereby eliminating the impact of dimensional differences in each indicator; through standardization, it is possible to avoid certain indicators from dominating the calculation due to their large original value range, thereby achieving fair comparison of each indicator; the initial weight coefficient is calculated based on the standardized importance index, which ensures that the sum of all weight coefficients is 1, thereby ensuring that the total weight is constant; by calculating the initial weight based on the standardized importance index, it is possible to ensure that the actual importance of each indicator is fully reflected in the total weight, thereby improving the scientificity and rationality of the model.

[0118] The design weight adjustment strategy specifically includes:

[0119] The design weight adjustment formula is: where ω i is the weight of the i-th element in the importance set, α i is the initial weight coefficient of the i-th element, I i is the importance score of the ith element, is the weighted sum of all elements and is used to normalize the weights so that the sum of all weights is 1.

[0120] By designing a weight adjustment formula, the weights can be dynamically adjusted according to the importance scores of each dimension, making the model more flexible and adaptable. The use of a formulaic calculation method can reduce the impact of human factors on weight setting. The formula uses an exponential function and normalization processing to standardize the weights to a value between 0 and 1, ensuring that the sum of all weights is 1. This standardization process makes the weight setting more standardized. By introducing importance scores, the formula can fully consider the differences in importance of different dimensions in risk assessment.

[0121] The user comprehensive scoring method specifically includes:

[0122] The weighted sum of each element in the importance set is: S = ∑ i ω i ×Score i where ω i The weight of the i-th element, Score i Get the importance score for the i-th element.

[0123] The comprehensive score S calculated by weighting the elements in the importance set provides a comprehensive and quantitative result for the evaluation model. In this way, the influence of each importance element can be comprehensively considered to obtain a unified scoring result. The weighted sum can fully reflect the role of each element in the overall evaluation and ensure the comprehensiveness and scientificity of the evaluation results. iThe weight of the i-th element determines the influence of the element in the comprehensive score; reasonable weight distribution can ensure that the contribution of each element in the total score matches its importance, avoiding some elements from being overly prominent or neglected.

[0124] The customer segmentation strategy specifically includes:

[0125] Divide user risk levels according to their comprehensive scores;

[0126] Very low risk: S ≥ 80;

[0127] Low risk: 70≦S<80;

[0128] Medium risk: 60≦S<70;

[0129] High risk: 50≦S<60;

[0130] Very high risk: S<50;

[0131] Set different risk level sets;

[0132] First, match the risk level according to the user's specific score, and then add the user to the corresponding risk level set.

[0133] Classifying users' risk levels according to their comprehensive scores is a key step in converting quantitative scoring results into actionable risk classifications. This method can help financial institutions quickly and accurately identify and classify users' credit risk status, thereby better managing loan approval and risk control processes. By clearly classifying risk levels, financial institutions can formulate corresponding strategies and measures for users of different risk levels to improve the effectiveness and accuracy of risk management. Setting different risk level sets and classifying various types of users into different categories such as very low risk, low risk, medium risk, high risk and very high risk can help financial institutions to systematically manage and classify users. First, matching the corresponding risk level according to the user's specific score, and then adding the user to the corresponding risk level set can ensure the accuracy and standardization of the risk assessment process.

[0134] It should be noted that, in this article, relational terms such as first and second, etc. are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "include", "comprise" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device.

[0135] The above is only a preferred embodiment of the present invention. It should be pointed out that for ordinary technicians in this technical field, several improvements and modifications can be made without departing from the technical principles of the present invention. These improvements and modifications should also be regarded as the scope of protection of the present invention.

Claims

1. A method for grouping overdue loan customers, characterized in that: include: Get the user's repayment behavior, denoted as I B , the repayment behavior includes the number of on-time repayments, total repayments and overdue payments in the last 12 months; Get basic user information, denoted as I D , the user's basic information includes annual income, age and occupation; Get the user's credit score, denoted as I C ; Get the user's loan amount, denoted as I A ; Get the user's loan purpose, denoted as I U ; Will I B ,I D ,I C ,I A ,I U Composing customer segment importance sets; Design a scoring formula for each element in the customer segmentation importance set; Design the initial weight coefficient of each element based on the importance of each element in the customer group importance set; Design weight adjustment strategies; The importance and weight of each element in the customer group importance set are integrated to obtain a comprehensive user scoring method; Use customer segmentation strategies to segment customers.

2. A method for grouping overdue loan customers according to claim 1, characterized in that: The scoring formula for each element in the design customer group importance set specifically includes: Repayment Behavior I B =2×R pay +3×(1-R default ); The R pay R is the ratio of the number of on-time repayments to the total number of repayments in the last 12 months. default The ratio of the expected number of repayments to the total number of repayments in the last 12 months; Basic InformationI D =0.5×I income +0.3×A age +0.2×P profession ; I income A is the user's annual income divided by the benchmark income. age P is the user's age divided by the base age. profession Rate careers by career stability and earning potential; User Credit Score I C =S credit ; The S credit For credit scores, credit scores are provided directly by credit scoring agencies, ranging from 300 to 850 divided by 850; User loan amount I A =0.7×A loan +0.3×(1-R income ); The A loan The loan amount is divided by the base amount, R income The ratio of the loan amount to annual income; User loan purpose I U =0.6×(1-R default )+0.4×U purpose ; The R default is the ratio of overdue payments to total repayments in the last 12 months, U purpose Score the purpose of the loan.

3. A method for grouping overdue loan customers according to claim 2, characterized in that: The career stability and income potential scores include: According to occupational stability, occupations are divided into high-stability occupations, medium-stability occupations, and low-stability occupations. The elements in the high-stability occupation set are civil servants, teachers, medical professionals and business managers; The elements in the medium stability occupation set are financial analysts, engineers and technical support; The elements in the low-stability occupation set are freelancers, sales representatives, contract workers and temporary workers; According to income potential, occupations are divided into a high-income potential occupation set, a medium-income potential occupation set and a low-income potential occupation set; The elements of the high-income potential occupation set are senior managers, doctors, financial industry executives, technology industry executives and top engineers; The elements of the medium-income potential occupation set are middle-level managers, lawyers, and senior technicians; The elements of the low-income potential occupation set are administrative assistants, service industry workers, and sales assistants; Let high stability occupations have a score of 3, medium stability occupations have a score of 2, and low stability occupations have a score of 1; Assume that the score of high-income potential occupations is 3, the score of medium-income potential occupations is 2, and the score of low-income potential occupations is 1; Let the stability weight be A, the income weight be B, and A+B=1; Then the career score = (stability score × A) + (income potential score × B).

4. A method for grouping overdue loan customers according to claim 2, characterized in that: The loan purpose score specifically includes: Set a loan purpose set, the elements in the set are consumer loan C1, housing loan C2, car loan C3, education loan C4, business loan C5, agricultural loan C6, investment loan C7 and emergency loan C8; The scoring formula for loan purpose is Score(C i )=w1×Risk(C i )+w2×Amount(C i )+w3×Repayment(Ci), where C i represents the i-th element in the loan purpose set, Risk(C i ) indicates category C i The risk level, Amount(C i ) represents the loan amount, Repayment(Ci) represents the repayment ability, w1, w2, w3 represent the weights of risk, amount, and repayment ability respectively, and w1+w2+w3=1.

5. A method for grouping overdue loan customers according to claim 1, characterized in that: The initial weight coefficients of the design elements specifically include: The importance index of each element is standardized to between 0 and 1, then Among them I i is the importance index of the i-th element, is the importance after normalization; Initial weight coefficient α i It can be calculated based on the standardized importance index, This ensures that all initial weight coefficients sum to 1.

6. A method for grouping overdue loan customers according to claim 1, characterized in that: The design weight adjustment strategy specifically includes: The design weight adjustment formula is: where ω i is the weight of the i-th element in the importance set, α i is the initial weight coefficient of the i-th element, I i is the importance score of the ith element, is the weighted sum of all elements and is used to normalize the weights so that the sum of all weights is 1.

7. A method for grouping overdue loan customers according to claim 1, characterized in that: The user comprehensive scoring method specifically includes: The weighted sum of each element in the importance set is: S = ∑ i ω i ×Score i where ω i The weight of the i-th element, Score i Get the importance score for the i-th element.

8. A method for grouping overdue loan customers according to claim 1, characterized in that: The customer segmentation strategy specifically includes: Divide user risk levels according to their comprehensive scores; Very low risk: S ≥ 80; Low risk: 70≦S<80; Medium risk: 60≦S<70; High risk: 50≦S<60; Very high risk: S<50; Set different risk level sets; First, match the risk level according to the user's specific score, and then add the user to the corresponding risk level set.

Images