CN116523627A - Credit line prediction method, device, equipment, medium and product - Google Patents

Abstract

The application provides a credit limit prediction method, a device, equipment, a medium and a product, and relates to the technical fields of big data and artificial intelligence, wherein the method comprises the following steps: acquiring client data of a client to be trusted, wherein the client data comprises enterprise internal financial data and enterprise external management data; predicting a financing constraint state of the customer to be trusted based on the customer data and a preset financing constraint model, wherein the financing constraint state is used for indicating whether the customer to be trusted is subjected to financing constraint; and carrying out credit limit prediction on the client to be credit limit based on the financing constraint state, the client data and a preset credit limit model to obtain a predicted credit limit. By determining the financing constraint state of the client to be trusted and combining with the client data, the credit line of the small and micro enterprise client is given more intuitively and accurately, and the accuracy of the credit line is improved.

Description

Credit line prediction method, device, equipment, medium and product

Technical Field

The application relates to the technical field of big data and artificial intelligence, in particular to a credit limit prediction method, a device, equipment, a medium and a product.

Background

In order to excite economic viability, financial institutions enact and issue a series of regulations and policies to support the development of small micro-enterprises, and the main means is to relieve the financing constraint of the small micro-enterprises by improving the financing environment of the small micro-enterprises through the implementation of general finance. However, in the credit risk field of general finance, small micro enterprises are limited by a plurality of factors such as imperfect management and credit records, serious information asymmetry and the like; on the other hand, the subjectivity of the financing demand is judged to be larger by the communication between the auditor and the small and micro-enterprise lending main body, and the problem that the credit line cannot meet the financing demand possibly occurs. That is, the present credit limit prediction method has low prediction accuracy when performing credit limit prediction on small micro enterprises.

Disclosure of Invention

The credit limit prediction method, the device, the equipment, the medium and the product can improve the accuracy of credit limit prediction.

In a first aspect, an embodiment of the present application provides a credit line prediction method, where the method includes:

acquiring client data of a client to be trusted, wherein the client data comprises enterprise internal financial data and enterprise external management data;

Predicting a financing constraint state of the customer to be trusted based on the customer data and a preset financing constraint model, wherein the financing constraint state is used for indicating whether the customer to be trusted is subjected to financing constraint;

and carrying out credit limit prediction on the client to be credit limit based on the financing constraint state, the client data and a preset credit limit model to obtain a predicted credit limit.

In a second aspect, the present application provides a credit line prediction apparatus, including:

the system comprises an acquisition module, a verification module and a verification module, wherein the acquisition module is used for acquiring client data of a client to be trusted, and the client data comprises enterprise internal financial data and enterprise external management data;

the first prediction module is used for predicting the financing constraint state of the client to be trusted based on the client data and a preset financing constraint model, wherein the financing constraint state is used for indicating whether the client to be trusted is subjected to financing constraint;

and the second prediction module is used for predicting the credit limit of the client to be trusted based on the financing constraint state, the client data and a preset credit limit model to obtain a predicted credit limit.

In a third aspect, an embodiment of the present application provides an electronic device, including: a processor and a memory storing computer program instructions;

The processor executes the computer program instructions to implement the credit line prediction method as in any one of the embodiments of the first aspect.

In a fourth aspect, an embodiment of the present application provides a computer storage medium, where computer program instructions are stored, where the computer program instructions implement a credit line prediction method as in any one of the embodiments of the first aspect when executed by a processor.

In a fifth aspect, embodiments of the present application provide a computer program product, where instructions in the computer program product, when executed by a processor of an electronic device, cause the electronic device to perform a credit line prediction method implemented as in any one of the embodiments of the first aspect.

The credit limit prediction method, device, equipment, medium and product in the embodiment of the application, wherein the method comprises the following steps: acquiring client data of a client to be trusted, wherein the client data comprises enterprise internal financial data and enterprise external management data; predicting a financing constraint state of the customer to be trusted based on the customer data and a preset financing constraint model, wherein the financing constraint state is used for indicating whether the customer to be trusted is subjected to financing constraint; and carrying out credit limit prediction on the client to be credit limit based on the financing constraint state, the client data and a preset credit limit model to obtain a predicted credit limit. By determining the financing constraint state of the client to be trusted and combining with the client data, the credit line of the small and micro enterprise client is given more intuitively and accurately, so that the accuracy of credit line prediction is improved.

Drawings

In order to more clearly illustrate the technical solutions of the embodiments of the present application, the drawings that are needed in the embodiments of the present application will be briefly described, and it is possible for a person skilled in the art to obtain other drawings according to these drawings without inventive effort.

FIG. 1 is a flowchart illustrating a credit limit prediction method according to an embodiment of the present disclosure;

FIG. 2 is a general flow chart of a credit prediction method according to an embodiment of the present application;

fig. 3 is a schematic structural diagram of a credit line prediction device according to an embodiment of the present application;

fig. 4 is a schematic structural diagram of an electronic device according to an embodiment of the present application.

Detailed Description

In order that the above objects, features and advantages of the present disclosure may be more clearly understood, a further description of aspects of the present disclosure will be provided below. It should be noted that, without conflict, the embodiments of the present disclosure and features in the embodiments may be combined with each other. The data acquisition, storage, use, processing and the like in the technical scheme meet the relevant regulations of national laws and regulations.

In the following description, numerous specific details are set forth in order to provide a thorough understanding of the present disclosure, but the present disclosure may be practiced otherwise than as described herein; it will be apparent that the embodiments in the specification are only some, but not all, embodiments of the disclosure.

It should be noted that in this document, relational terms such as "first" and "second" and the like are used solely to distinguish one entity or action from another entity or action without necessarily requiring or implying any actual such relationship or order between such entities or actions. Moreover, the terms "comprises," "comprising," or any other variation thereof, are intended to cover a non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements does not include only those elements but may include other elements not expressly listed or inherent to such process, method, article, or apparatus. Without further limitation, an element defined by the phrase "comprising one … …" does not exclude the presence of other like elements in a process, method, article, or apparatus that comprises an element.

In order to solve the problems in the prior art, the embodiment of the application provides a credit limit prediction method, a credit limit prediction device, credit limit prediction equipment, a credit limit prediction medium and a credit limit prediction product. The credit limit prediction method provided by the embodiment of the application is first described below.

Fig. 1 is a flow chart illustrating a credit limit prediction method according to an embodiment of the present application. As shown in fig. 1, the method specifically may include the following steps:

Step 101, obtaining client data of a client to be trusted, wherein the client data comprises enterprise internal financial data and enterprise external operation data;

the method and the device are mainly applied to credit line prediction of the small micro enterprises by the financial institutions after the small micro enterprises initiate credit request. The client to be trusted is the small micro enterprise client which initiates the trust request. The intra-enterprise financial data may be customer support class data and capital funds class data, such as: the amount of expenditure, the amount of income, annual profit, historical loans, etc. The enterprise external management data comprises basic data and credit indexes, and specifically can be related data acquired from clients through an external industrial and commercial information platform and a credit investigation system, for example: one or more of registered capital, registered time, business scope, business terms, stakeholder information, high management information, liability information, and poor information.

Step 102, predicting a financing constraint state of the customer to be trusted based on the customer data and a preset financing constraint model, wherein the financing constraint state is used for indicating whether the customer to be trusted is subjected to financing constraint;

the financing constraint means that the target credit line of the client to be trusted is larger than the actual credit line, i.e. the actual credit line does not reach the target credit line. The financing constraint model is mainly used for predicting whether a client to be trusted will be subjected to financing constraint. When the prediction is carried out, the client data can be input into the financing constraint model, and the predicted financing constraint state of the client to be trusted is directly output.

And step 103, carrying out credit limit prediction on the client to be trusted based on the financing constraint state, the client data and a preset credit limit model to obtain a predicted credit limit.

The financing constraint state comprises financing constraints and non-financing constraints, the financing constraints mean that the clients to be trusted are subjected to the financing constraints, and the non-financing constraints mean that the clients to be trusted are not subjected to the financing constraints. The credit limit model is used for predicting the credit limit of the client to be trusted.

In this embodiment, client data of a client to be trusted is obtained, where the client data includes enterprise internal financial data and enterprise external operation data; predicting a financing constraint state of the customer to be trusted based on the customer data and a preset financing constraint model, wherein the financing constraint state is used for indicating whether the customer to be trusted is subjected to financing constraint; and carrying out credit limit prediction on the client to be credit limit based on the financing constraint state, the client data and a preset credit limit model to obtain a predicted credit limit. Because the existing credit prediction method is seldom considering the effect of financing constraint on the credit granting of the small and micro enterprise clients, the credit granting of the small and micro enterprise clients is more intuitively and accurately given out by determining the financing constraint state of the clients to be trusted and combining with the client data, and the accuracy of credit prediction is improved.

In an embodiment of the present application, the step of predicting the credit line of the client to be trusted based on the financing constraint state, the client data and a preset credit line model to obtain a preset credit line includes:

under the condition that the financing constraint state of the client to be trusted is financing constraint, carrying out credit line prediction on the client to be trusted according to the internal financial data, the enterprise external operation data and a preset credit line model to obtain predicted credit line, wherein the internal financial data or the enterprise external operation data comprise financing constraint agent variables;

under the condition that the financing constraint state of the client to be trusted is non-financing constraint, determining the rest data except for preset financing constraint agent variables in the enterprise internal financial data and the enterprise external operation data; and

and carrying out credit line prediction on the client to be credit-granted according to the rest data and a preset credit line model to obtain predicted credit line.

Since the financing constraint state is an unobservable or unmeasurable variable, in this embodiment the financing constraint proxy variable is used to represent the financing constraint state. In this embodiment, the financing constraint proxy variable refers to an index that can reflect the financing constraint state of the client to be trusted, and is contained in the internal financial data and the external business data, i.e. can be represented by one of the internal financial data and the external business data, for example: one of the base data, customer support class, capital fund class, and credit class indicators. Of course, it is also possible to express a combination of a plurality of data therein, for example: any combination of a plurality of data in the base data, the customer support class, the capital funding class, and the credit class indicator.

Under the condition that the financing constraint state of the client to be trusted is the financing constraint, because the financing constraint agent variable is expressed as the financing constraint state of the client to be trusted, the enterprise internal financial data and the enterprise external operation data comprising the financing constraint agent variable are required to be input into a credit line model, and the credit line of the client to be trusted is predicted, so that the predicted credit line is obtained.

Under the condition that the financing constraint state of the client to be trusted is non-financing constraint, the financing constraint agent variable is expressed as the financing constraint state of the client to be trusted, and the client to be trusted is non-financing constraint state, so that the financing constraint agent variable in the internal financial data and the external operation data needs to be removed, and then the internal financial data and the external operation data after the financing constraint agent variable is removed are input into a credit line model to output a predicted credit line. The remaining data in this embodiment is the internal financial data and the external business data after the financing constraint agent variables are removed.

In this embodiment, through the above steps, the financing constraint proxy variable representing the financing constraint state of the client to be trusted may be respectively matched with the financing constraint state of the client to be trusted, and when the credit limit prediction is performed, the influence of the financing constraint state on the credit limit of the client to be trusted is fully considered, and meanwhile, the credit limit is comprehensively predicted through the repayment capability and the credit level reflected by the internal financial data and the external management data, so that the accuracy of the credit limit prediction is improved.

In an embodiment of the present application, before the step of obtaining the client data of the client to be trusted, the method further includes:

selecting a plurality of first customer samples;

determining first sample data for a plurality of the first customer samples, the first sample data comprising first internal financial data and first external business data;

labeling the financing constraint state of each first customer sample according to a preset labeling rule;

training a first basic model by taking a plurality of first customer samples as training samples to obtain the financing constraint model;

the first customer sample may be randomly selected from a database of a financial institution, referring to customers who have completed credit, including both samples subject to financing constraints and samples not subject to financing constraints. The first internal financial data and the first external operation data are identical to the types of the internal financial data and the external financial data in the above embodiments, and are not described herein.

When labeling the first customer samples, each first customer sample is labeled as either a financing-constrained customer sample or a non-financing-constrained customer sample based on credit information for the first customer sample information.

The first base model may be an xgboost model (extreegradientboosting) of multiple trees. When model training is carried out, the first customer sample marked with the first internal financial data and the first external operation data are used as training data, the marked financing constraint state is used as a training label, and the financing constraint model is obtained through training. And the final output result of the financing constraint model obtained by training is the financing constraint state of the predicted client to be trusted.

By training the financing constraint model, the prediction of the financing constraint state of the client to be trusted is realized, and the state result fully considers the financing requirement of the client to be trusted, so that the prediction of the credit limit is more accurate.

In another embodiment, after the step of training the first base model using the plurality of first customer samples as training samples to obtain the financing constraint model, the method further includes:

and determining the financing constraint proxy variable in the first internal financial data and the first external business data according to the financing constraint model.

The financing constraint agent variable is a significant characteristic variable of the financing constraint model in the training process of the financing constraint model, and after the financing constraint model is trained, the outside-bag data (OOB data) is used for carrying out quantitative calculation of the characteristic importance. And calculating the characteristic change rate of each data in the first internal financial data and the first external operation data, and finally quantifying the importance of each characteristic in the first internal financial data and the first external operation data according to the change rate sequence. And selecting one or more features with the forefront importance as financing constraint agent variables.

In this embodiment, by mining financing constraint proxy variables and using the same as an important supplementary quantization factor for evaluating credit limits of clients to be trusted, training of a subsequent credit limit model is more accurate.

In yet another embodiment of the present application, the step of labeling the financing constraint status of each of the first customer samples according to a preset labeling rule includes:

obtaining credit information for each of the first customer samples;

marking the first customer sample as a non-financing constraint customer sample when the credit information meets a repayment condition and the credit information meets an unoverdue condition;

and marking the first customer sample as a financing constraint customer sample in the case that the credit information does not meet a repayment condition or the credit information is overdue.

Credit information refers to credit products used by customers, the status of the continued credit within 1-2 months before expiration of an year, and the status of the expiration of the credit products.

And when the credit information meets the repayment condition and the credit information is not overdue, namely that the client performs repayment or postpones within 1-2 months before expiration of one year, and when the credit information is not overdue, the credit line of the client is indicated to not meet the credit line of the requirement, namely, the first client sample is marked as a financing constraint client sample.

And under the condition that the credit information does not meet the repayment condition or the credit information is overdue, namely that the client does not carry out repayment or delay within 1-2 months before the annual expiration, or that the credit information is overdue, the credit line of the client is indicated to meet the credit line of the client, namely, the first client sample is marked as a non-financing constraint client sample.

In this embodiment, the financing constraint state of the first customer sample is marked, so that the financing constraint model obtained by training is more accurate, and meanwhile, the applicability and non-limitation of the first customer sample and the range thereof are ensured through the calibration rule.

In an embodiment of the present application, after the step of determining the financing constraint proxy variable in the first sample data based on the financing constraint model, the method further comprises:

obtaining a plurality of second client samples, wherein the second client samples are clients who finish credit giving;

determining second sample data and historical credit limits for a plurality of the second customer samples, the sample data including second internal financial data and second external business data;

determining financing constraint states of a plurality of second client samples according to the financing constraint model;

Labeling the financing constraint state and the historical credit line of each second client sample;

and training the second basic model by taking a plurality of second client samples as training samples to obtain the credit limit model.

The second customer sample may be randomly selected from a database of the financial institution, referring to customers who have completed credit authorization. The first internal financial data and the first external operation data are identical to the types of the internal financial data and the external financial data in the above embodiments, and are not described herein. Since the second client sample is a client who has completed credit, the historical credit limit of the second client sample can be determined directly in the database of the financial institution.

And determining the financing constraint state of each second client sample according to the financing constraint model obtained through training. When the credit line model is trained, the financing constraint state and the historical credit line are marked on the corresponding second client sample. And then training the second basic model by using the second internal financial data and the second external operation data as training data and the historical credit line data as training results to obtain a trained credit line model, wherein the second basic model is a lightgbm model (LightGradient BoostingMachine). The final output result of the credit limit model obtained through training is the predicted credit limit of the client to be trusted.

In this embodiment, the credit line model is obtained through training, so as to realize the credit line prediction of the clients to be trusted, the basis line model is trained by using a lightgbm algorithm based on historical credit data, and the basis line of the clients of the small micro enterprises is more intuitively and accurately given by combining two mechanisms of repayment capability and financing constraint, so as to evaluate the credit line of the clients for enterprise operation.

In an embodiment of the present application, after the step of predicting the credit line of the client to be trusted and obtaining the predicted credit line based on the financing constraint state, the client data and a preset credit line model, the method further includes:

obtaining the default probability of the client to be trusted;

determining a risk level of the client to be trusted according to the default probability;

and adjusting the predicted credit limit according to the risk coefficient corresponding to the risk level to obtain the credit limit of the client to be trusted.

When the credit line of the client to be trusted is predicted through the credit line model, the repayment risk level of the client to be trusted is not considered, so that when the final credit line of the client to be trusted is determined, the predicted credit line needs to be further adjusted.

The probability of breach of the client to be trusted can be calculated by historical credit information of the client to be trusted, for example, the probability of breach of the client to be trusted is determined by calculating the ratio between the number of breach of the client to be trusted and the total loan number.

After determining the default probability of the client to be trusted, dividing the grades according to the default probability of the client, assigning a risk coefficient to each grade, and weighting the adjustment coefficient on the predicted credit limit. When the risk coefficient is smaller, the credit line obtained by the user is higher or the opportunity for obtaining the line by the user is larger; and when the risk coefficient reaches the standard value, the credit limit is not given to the user.

In this embodiment, by comprehensively considering the financing demand, repayment capability and risk level of the customer, the credit line of the customer is evaluated, so that balance between the repayment capability and the financing demand constraint of the customer is achieved, and the credit risk of the financial institution is reduced.

The following illustrates a credit line prediction method provided in the present application, and fig. 2 shows an overall flow chart of the credit line prediction method provided in the present application. The credit limit prediction method specifically comprises the following steps:

(1) A sample defining a target financing constraint (i.e., a first customer sample and a second customer sample);

(2) Selecting internal data and external data (namely internal financial data and external business data) as alternative proxy variables of financing constraint, wherein the alternative proxy variables mainly comprise basic data, customer support class, capital fund class and credit class indexes;

(3) Training a financing constraint model and a credit line model;

(1) the training method for training the financing constraint model specifically comprises the following steps:

a. randomly selecting n customer samples (i.e., a first customer sample);

b. labeling financing constraint states of the n customers according to the definition that the repayment or delay is satisfied and no overdue occurs;

c. through marked customer samples, using internal data and external data (namely first internal financial data and first external business data) as training data, marking data (namely financing constraint state) as training results, and training a multi-tree xgboost model (namely a first basic model);

the external data is related data acquired from the clients through an external industrial and commercial information platform and a credit investigation system, and comprises the following steps: one or more of registered capital, registered time, business scope, business terms, stakeholder information, high management information, liability information, and poor information. The internal data refers to customer support class and capital funds class data.

d. The final output result of the trained xgboost model is the financing constraint effect of the prediction client, and the feature importance of the financing constraint agent variable is judged.

(2) Training a credit limit model, wherein the training method specifically comprises the following steps:

a. randomly selecting m customer samples (i.e., a second customer sample);

b. labeling financing constraint states of the m clients according to the financing constraint model;

c. through marked real credit client sample credit line data (i.e. historical credit line), internal data and external data (i.e. second internal financial data and second external business data) are used as training data, the credit line data is used as training results, and a lightgbm model (i.e. historical credit line) is trained;

the external data is basic data through external business information. The internal data refers to repayment capability, historical loan, and financing constraint data.

d. Inputting the internal data and the external data into a trained lightgbm model; if m clients are marked as financing constraint clients through the financing constraint model, the internal data and the external data comprise proxy variable data of the financing constraint class; otherwise, not including;

e. the final output result of the trained lightgbm model is the predicted client's basic credit line (i.e., predicted credit line).

(4) And adjusting the credit line, layering the clients according to the default probability distribution of the clients, giving a risk coefficient to the clients of each level, and weighting the risk coefficient to the basic line to obtain the final credit line.

Based on the algorithm flow, in practical application, the agent variable of the financing constraint of the client is mined through the client characteristics of the internal and external data, the financing constraint, repayment capability, risk level and historical lending condition of the client are comprehensively considered, and the final credit giving amount is obtained in a quantified mode.

The following will exemplify a loan to a bank by the customer a.

First, the bank establishes a financing constraint model.

Firstly, randomly selecting a plurality of N first customer samples in a bank database, and acquiring information such as expenditure amount, income amount, annual profit, historical lending, registered capital, registered time, business scope, operation period, stakeholder information, high-management information, liability information, bad information and the like of the first customer samples;

secondly, marking the customers which meet the repayment condition in the N first customer samples and the credit information meet the non-overdue as financing constraint customer samples, and marking the customers which do not meet the repayment condition or the credit information overdue as non-financing constraint customer samples;

And finally, training the xgboost model by taking N first client samples as training data, outputting a financing constraint model and financing constraint proxy variables.

And secondly, the bank establishes a limit prediction model.

Firstly, randomly selecting M second customer samples with credit authorization completion in a bank database, and obtaining information such as expenditure amount, income amount, annual profit, historical loan, registered capital, registration time, business scope, business deadline, stakeholder information, high management information, liability information, bad information and the like of the second customer samples, and obtaining specific credit authorization of the second customer samples;

secondly, predicting financing constraint states of M second client samples through a financing constraint model, and marking specific credit limit and the financing constraint states of the second client samples;

and finally, taking the second client sample as training data, and inputting the training data into the lightgbm model for training to obtain the credit limit model.

Thirdly, the bank predicts the credit limit of the client A.

Firstly, obtaining information such as expenditure amount, income amount, annual profit, historical borrowing, registered capital, registered time, business scope, business period, stakeholder information, high management information, liability information, bad information and the like of a client A;

Secondly, inputting the information into a financing constraint model to obtain a financing constraint state of the client A;

finally, under the condition that the financing constraint state of the client A is the financing constraint, information such as expenditure amount, income amount, annual profit, history lending, registered capital, registration time, business scope, operation period, stockholder information, high management information, liability information, bad information and the like of the client A are input into a credit line model to obtain predicted credit line;

and under the condition that the financing constraint state of the client A is not the financing constraint, inputting the rest data except the financing constraint agent variable in the expenditure amount, the income amount, the annual profit, the historical loan, the registered capital, the registered time, the business scope, the business period, the stockholder information, the high management information, the liability information and the bad information of the client A into a credit line model to obtain the predicted credit line.

And fourthly, determining the credit limit of the client A.

Firstly, obtaining the default probability of a client A based on the default times and the total loan times in the history credit information of the client A;

secondly, determining the risk level of the default probability of the client A at the bank and the corresponding risk coefficient;

And finally, obtaining the product of the risk coefficient and the predicted credit limit, and taking the product as the credit limit of the client A.

The credit limit prediction method provided by the application has the following beneficial effects:

(1) Improving the credit limit accuracy of the clients to be trusted;

according to the method, under the condition of comprehensively considering the existing internal and external data, the financing constraint model is used for analyzing the financing demands of the clients, the decision tree algorithm is used for screening agent variable characteristics of the financing constraint, and the obvious financing constraint agent variable characteristics are incorporated into the credit line model, so that the credit line model can evaluate the credit line of the clients under the background of reducing subjective factors and comprehensively considering the financing demands, repayment capacity and risk level of the clients, and can give out relevant factors for judging the differentiated credit line, and the influence and result generated by the constraint balance of the repayment capacity and the financing demands of the clients are reflected.

(2) Reducing the credit risk and improving the viscosity of the client;

by accurately evaluating the credit line of the client to be trusted and combining the risk level of the client to be trusted, the credit line of the client is adjusted, so that the credit risk of the client with high default probability can be effectively reduced, the client can be promoted to keep good credit level, and the viscosity of the client is further improved.

Fig. 3 is a schematic structural diagram of a credit line predicting device according to an embodiment of the present application, and for convenience of explanation, only a portion relevant to the embodiment of the present application is shown.

Referring to fig. 3, the credit limit prediction apparatus 300 may include:

a first obtaining module 301, configured to obtain client data of a client to be trusted, where the client data includes internal financial data of an enterprise and external business data of the enterprise;

the first prediction module 302 is configured to predict a financing constraint state of the client to be trusted based on the client data and a preset financing constraint model, where the financing constraint state is used to indicate whether the client to be trusted is subjected to financing constraint;

and the second prediction module 303 is configured to predict the credit limit of the client to be trusted based on the financing constraint state, the client data and a preset credit limit model, and obtain a predicted credit limit.

Optionally, the second prediction module 303 includes:

the first prediction unit is used for predicting the credit line of the client to be trusted according to the internal financial data, the enterprise external operation data and a preset credit line model under the condition that the financing constraint state of the client to be trusted is the financing constraint, so as to obtain a predicted credit line, wherein the internal financial data or the enterprise external operation data comprise financing constraint agent variables;

The second prediction unit is used for determining the rest data except the preset financing constraint agent variable in the enterprise internal financial data and the enterprise external operation data under the condition that the financing constraint state of the client to be trusted is non-financing constraint; and

and carrying out credit line prediction on the client to be credit-granted according to the rest data and a preset credit line model to obtain predicted credit line.

Optionally, the credit limit prediction device 300 further includes:

a first selecting module for selecting a plurality of first customer samples;

a first determination module for determining first sample data for a plurality of the first customer samples, the first sample data comprising first internal financial data and first external business data;

the first labeling module is used for labeling the financing constraint state of each first customer sample according to a preset labeling rule;

and the first training module is used for training the first basic model by taking a plurality of first client samples as training samples to obtain the financing constraint model.

Optionally, the first labeling module includes:

an acquisition unit configured to acquire credit information of each of the first customer samples;

The first labeling unit is used for labeling the first client sample as a financing constraint client sample under the condition that the credit information meets the repayment condition and the credit information meets the condition of not overdue;

and the second labeling unit is used for labeling the first client sample as a non-financing constraint client sample under the condition that the credit information does not meet the repayment condition or the credit information is overdue.

Optionally, the credit limit prediction device 200 further includes:

the second selecting module is used for acquiring a plurality of second client samples, wherein the second client samples are clients who finish credit giving;

a second determining module, configured to determine second sample data and historical credit limits of a plurality of second customer samples, where the sample data includes second internal financial data and second external business data;

the third determining module is used for determining financing constraint states of a plurality of second client samples according to the financing constraint model;

the second labeling module is used for labeling the financing constraint state and the historical credit limit of each second client sample;

and the second training module is used for training the second basic model by taking a plurality of second client samples as training samples to acquire the credit limit model.

Optionally, the credit limit prediction device 300 further includes:

the second acquisition module is used for acquiring the default probability of the client to be trusted;

the judging module is used for determining the risk level of the client to be trusted according to the default probability;

and the adjustment module is used for adjusting the predicted credit line according to the risk coefficient corresponding to the risk level to obtain the credit line of the client to be trusted.

The credit limit prediction device 300 provided in this embodiment of the present application can implement each process implemented by the foregoing method embodiment, and in order to avoid repetition, a description is omitted here.

It will be apparent to those skilled in the art that, for convenience and brevity of description, only the above-described division of the functional units and modules is illustrated, and in practical application, the above-described functional distribution may be performed by different functional units and modules according to needs, i.e. the internal structure of the apparatus is divided into different functional units or modules to perform all or part of the above-described functions. The functional units and modules in the embodiment may be integrated in one processing unit, or each unit may exist alone physically, or two or more units may be integrated in one unit, where the integrated units may be implemented in a form of hardware or a form of a software functional unit. In addition, specific names of the functional units and modules are only for convenience of distinguishing from each other, and are not used for limiting the protection scope of the present application. The specific working process of the units and modules in the above system may refer to the corresponding process in the foregoing method embodiment, which is not described herein again.

Fig. 3 shows a schematic hardware structure of an electronic device according to an embodiment of the present application.

The device may include a processor 401 and a memory 402 in which program instructions are stored.

The steps of any of the various method embodiments described above are implemented when the processor 401 executes a program.

For example, a program may be divided into one or more modules/units, which are stored in the memory 402 and executed by the processor 401 to complete the present application. One or more of the modules/units may be a series of program instruction segments capable of performing specific functions to describe the execution of the program in the device.

In particular, the processor 401 described above may include a Central Processing Unit (CPU), or a specific integrated circuit (ApplicationSpecificIntegratedCircuit, ASIC), or may be configured to implement one or more integrated circuits of embodiments of the present application.

Memory 402 may include mass storage for data or instructions. By way of example, and not limitation, memory 402 may comprise a hard disk drive (HardDiskDrive, HDD), a floppy disk drive, flash memory, an optical disk, a magneto-optical disk, a magnetic tape, or a universal serial bus (UniversalSerialBus, USB) drive, or a combination of two or more of these. Memory 402 may include removable or non-removable (or fixed) media, where appropriate. Memory 402 may be internal or external to the integrated gateway disaster recovery device, where appropriate. In a particular embodiment, the memory 402 is a non-volatile solid state memory.

The memory may include Read Only Memory (ROM), random Access Memory (RAM), magnetic disk storage media devices, optical storage media devices, flash memory devices, electrical, optical, or other physical/tangible memory storage devices. Thus, in general, the memory includes one or more tangible (non-transitory) readable storage media (e.g., memory devices) encoded with software comprising computer-executable instructions and when the software is executed (e.g., by one or more processors) it is operable to perform the operations described with reference to methods in accordance with aspects of the present disclosure.

The processor 401 implements any of the methods of the above embodiments by reading and executing program instructions stored in the memory 402.

In one example, the electronic device may also include a communication interface 403 and a bus 410. The processor 401, the memory 402, and the communication interface 403 are connected to each other by a bus 410 and perform communication with each other.

The communication interface 403 is mainly used to implement communication between each module, device, unit and/or apparatus in the embodiments of the present application.

Bus 410 includes hardware, software, or both, coupling components of the online data flow billing device to each other. By way of example, and not limitation, the buses may include an Accelerated Graphics Port (AGP) or other graphics bus, an Enhanced Industry Standard Architecture (EISA) bus, a Front Side Bus (FSB), a HyperTransport (HT) interconnect, an Industry Standard Architecture (ISA) bus, an infiniband interconnect, a Low Pin Count (LPC) bus, a memory bus, a micro channel architecture (MCa) bus, a Peripheral Component Interconnect (PCI) bus, a PCI-Express (PCI-X) bus, a Serial Advanced Technology Attachment (SATA) bus, a video electronics standards association local (VLB) bus, or other suitable bus, or a combination of two or more of the above. Bus 410 may include one or more buses, where appropriate. Although embodiments of the present application describe and illustrate a particular bus, the present application contemplates any suitable bus or interconnect.

In addition, in combination with the method in the above embodiment, the embodiment of the application may be implemented by providing a storage medium. The storage medium has program instructions stored thereon; the program instructions, when executed by a processor, implement any of the methods of the embodiments described above.

The embodiment of the application further provides a chip, the chip includes a processor and a communication interface, the communication interface is coupled with the processor, the processor is used for running a program or instructions, the processes of the above method embodiment are realized, the same technical effects can be achieved, and in order to avoid repetition, the description is omitted here.

It should be understood that the chips referred to in the embodiments of the present application may also be referred to as system-on-chip chips, chip systems, or system-on-chip chips, etc.

The embodiments of the present application provide a computer program product, which is stored in a storage medium, and the program product is executed by at least one processor to implement the respective processes of the above method embodiments, and achieve the same technical effects, and are not repeated herein.

It should be clear that the present application is not limited to the particular arrangements and processes described above and illustrated in the drawings. For the sake of brevity, a detailed description of known methods is omitted here. In the above embodiments, several specific steps are described and shown as examples. However, the method processes of the present application are not limited to the specific steps described and illustrated, and those skilled in the art can make various changes, modifications, and additions, or change the order between steps, after appreciating the spirit of the present application.

The functional blocks shown in the above block diagrams may be implemented in hardware, software, firmware, or a combination thereof. When implemented in hardware, it may be, for example, an electronic circuit, an Application Specific Integrated Circuit (ASIC), suitable firmware, a plug-in, a function card, or the like. When implemented in software, the elements of the present application are the programs or code segments used to perform the required tasks. The program or code segments may be stored in a machine readable medium or transmitted over transmission media or communication links by a data signal carried in a carrier wave. A "machine-readable medium" may include any medium that can store or transfer information. Examples of machine-readable media include electronic circuitry, semiconductor memory devices, ROM, flash memory, erasable ROM (EROM), floppy disks, CD-ROMs, optical disks, hard disks, fiber optic media, radio Frequency (RF) links, and the like. The code segments may be downloaded via computer grids such as the internet, intranets, etc.

It should also be noted that the exemplary embodiments mentioned in this application describe some methods or systems based on a series of steps or devices. However, the present application is not limited to the order of the above-described steps, that is, the steps may be performed in the order mentioned in the embodiments, may be different from the order in the embodiments, or several steps may be performed simultaneously.

Aspects of the present disclosure are described above with reference to flowchart illustrations and/or block diagrams of methods, apparatus (systems) and program products according to embodiments of the disclosure. It will be understood that each block of the flowchart illustrations and/or block diagrams, and combinations of blocks in the flowchart illustrations and/or block diagrams, can be implemented by computer program instructions. These program instructions may be provided to a processor of a general purpose computer, special purpose computer, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, enable the implementation of the functions/acts specified in the flowchart and/or block diagram block or blocks. Such a processor may be, but is not limited to being, a general purpose processor, a special purpose processor, an application specific processor, or a field programmable logic circuit. It will also be understood that each block of the block diagrams and/or flowchart illustration, and combinations of blocks in the block diagrams and/or flowchart illustration, can be implemented by special purpose hardware which performs the specified functions or acts, or combinations of special purpose hardware and computer instructions.

In the foregoing, only the specific embodiments of the present application are described, and it will be clearly understood by those skilled in the art that, for convenience and brevity of description, the specific working processes of the systems, modules and units described above may refer to the corresponding processes in the foregoing method embodiments, which are not repeated herein. It should be understood that the scope of the present application is not limited thereto, and any person skilled in the art can easily conceive various equivalent modifications or substitutions within the technical scope of the present application, which are intended to be included in the scope of the present application.

Claims (10)

1. A credit prediction method is characterized by comprising the following steps:

acquiring client data of a client to be trusted, wherein the client data comprises enterprise internal financial data and enterprise external management data;

predicting a financing constraint state of the customer to be trusted based on the customer data and a preset financing constraint model, wherein the financing constraint state is used for indicating whether the customer to be trusted is subjected to financing constraint;

and carrying out credit limit prediction on the client to be credit limit based on the financing constraint state, the client data and a preset credit limit model to obtain a predicted credit limit.

2. The credit line prediction method according to claim 1, wherein the step of predicting the credit line of the client to be trusted based on the financing constraint state, the client data and a preset credit line model to obtain a predicted credit line includes:

under the condition that the financing constraint state of the client to be trusted is financing constraint, carrying out credit line prediction on the client to be trusted according to the internal financial data, the enterprise external operation data and the credit line model to obtain predicted credit line, wherein the internal financial data or the enterprise external operation data comprise financing constraint agent variables;

Under the condition that the financing constraint state of the client to be trusted is non-financing constraint, determining the rest data except for preset financing constraint agent variables in the enterprise internal financial data and the enterprise external operation data; and

and carrying out credit limit prediction on the client to be credit-granted according to the rest data and the credit limit model to obtain predicted credit limit.

3. The credit prediction method according to claim 1, wherein before the step of acquiring the client data of the client to be trusted, the method further comprises:

selecting a plurality of first customer samples;

determining first sample data for a plurality of the first customer samples, the first sample data comprising first internal financial data and first external business data;

labeling the financing constraint state of each first customer sample according to a preset labeling rule;

and training the first basic model by taking a plurality of first customer samples as training samples to obtain the financing constraint model.

4. The credit prediction method according to claim 3, wherein the step of labeling the financing constraint status of each of the first customer samples according to a preset labeling rule includes:

Obtaining credit information for each of the first customer samples;

marking the first customer sample as a financing constraint customer sample when the credit information meets a repayment condition and the credit information meets an unoverdue condition;

marking the first customer sample as a non-financing constraint customer sample if the credit information does not satisfy a repayment condition or the credit information is overdue.

5. The credit prediction method of claim 3, wherein after the step of determining the financing constraint proxy variable in the first sample data based on the financing constraint model, the method further comprises:

obtaining a plurality of second client samples, wherein the second client samples are clients who finish credit giving;

determining second sample data and historical credit limits for a plurality of the second customer samples, the sample data including second internal financial data and second external business data;

determining financing constraint states of a plurality of second client samples according to the financing constraint model;

labeling the financing constraint state and the historical credit line of each second client sample;

and training the second basic model by taking a plurality of second client samples as training samples to obtain the credit limit model.

6. The credit line prediction method according to claim 1, wherein after the step of predicting the credit line of the client to be trusted and obtaining the predicted credit line based on the financing constraint state, the client data and a preset credit line model, the method further comprises:

obtaining the default probability of the client to be trusted;

determining a risk level of the client to be trusted according to the default probability;

and adjusting the predicted credit limit according to the risk coefficient corresponding to the risk level to obtain the credit limit of the client to be trusted.

7. A credit prediction device, characterized in that the device comprises:

the first acquisition module is used for acquiring client data of a client to be trusted, wherein the client data comprises enterprise internal financial data and enterprise external management data;

the first prediction module is used for predicting the financing constraint state of the client to be trusted based on the client data and a preset financing constraint model, wherein the financing constraint state is used for indicating whether the client to be trusted is subjected to financing constraint;

and the second prediction module is used for predicting the credit limit of the client to be trusted based on the financing constraint state, the client data and a preset credit limit model to obtain a predicted credit limit.

8. An electronic device, the device comprising: a processor and a memory storing computer program instructions;

the credit prediction method according to any one of claims 1-6 is implemented when the processor executes the computer program instructions.

9. A computer readable storage medium having stored thereon computer program instructions which when executed by a processor implement the credit line prediction method as claimed in any one of claims 1 to 6.

10. A computer program product, characterized in that instructions in the computer program product, when executed by a processor of an electronic device, cause the electronic device to perform the credit prediction method as claimed in any one of claims 1-6.