Credit risk determination method and device, storage medium and electronic equipment

Abstract

The invention discloses a credit risk determination method and device, a storage medium and electronic equipment. The method relates to the technical field of data processing, and comprises the following steps: carrying out credit risk assessment on target data information of a to-be-assessed target object through a target risk assessment model to obtain a risk prediction value; according to the risk prediction value, judging whether the target object has a first type of repayment default risk, and obtaining a first judgment result; and if the first judgment result represents that the target object does not have the first type of repayment default risk, judging whether the target object has a second type of repayment default risk according to the risk prediction value to obtain a second judgment result, and determining credit risk information corresponding to the target object according to the first judgment result and / or the second judgment result. Through the method and the device, the problem of relatively low accuracy of the credit risk caused by evaluating the credit risk of the borrower through a fixed risk inspection rule in related technologies is solved.

Description

Technical Field

[0001] The present application relates to the field of data processing technology, and in particular to a method and device for determining credit risk, a storage medium, and an electronic device. Background Art

[0002] Credit risk refers to the potential risk that the borrower cannot repay the principal and interest on time during the loan process. When lending, credit institutions need to assess the borrower's credit status and debt repayment ability to reduce credit risk. In existing technologies, risk testing rules are often formulated based on expert experience to assess the borrower's credit risk, but this method is often less accurate.

[0003] Currently, no effective solution has been proposed to address the problem that the credit risk of borrowers is assessed through fixed risk testing rules in related technologies, resulting in relatively low accuracy of credit risk. Summary of the invention

[0004] The main purpose of the present application is to provide a method and device for determining credit risk, a storage medium and an electronic device to solve the problem in the related art that the credit risk of a borrower is assessed through fixed risk verification rules, resulting in low accuracy of credit risk.

[0005] In order to achieve the above-mentioned purpose, according to one aspect of the present application, a method for determining credit risk is provided. The method comprises: performing credit risk assessment on target data information of a target object to be assessed through a target risk assessment model to obtain a risk prediction value, wherein the target data information at least includes the loan information and credit information of the target object; judging whether the target object has a first type of repayment default risk based on the risk prediction value to obtain a first judgment result; if the first judgment result indicates that the target object does not have the first type of repayment default risk, judging whether the target object has a second type of repayment default risk based on the risk prediction value to obtain a second judgment result, wherein the default time of the first type of repayment default risk and the second type of repayment default risk are different; determining the credit risk information corresponding to the target object based on the first judgment result and / or the second judgment result.

[0006] Furthermore, the target risk assessment model is obtained by adopting the following steps: obtaining a training sample set, wherein the training sample set includes multiple training samples and a true label corresponding to each training sample, each training sample is composed of historical loan information and historical credit information of the sample object, and the true label includes a first label and a second label, the first label is used to characterize whether the training sample has a first type of repayment default risk, and the second label is used to characterize whether the training sample has a second type of repayment default risk; training the initial risk assessment model based on the training sample set to obtain the target risk assessment model.

[0007] Furthermore, the initial risk assessment model is trained based on the training sample set to obtain the target risk assessment model, including: processing the training sample set through the initial risk assessment model to obtain a predicted value corresponding to each training sample; determining a target loss value based on the predicted value and the true label; and training the initial risk assessment model based on the target loss value to obtain the target risk assessment model.

[0008] Furthermore, determining the target loss value based on the predicted value and the true label includes: if the first label of the first training sample is a first preset value, determining the weight value corresponding to the first training sample based on the loan amount corresponding to the first training sample; if the first label of the second training sample is a second preset value, determining the second weight value corresponding to the second training sample based on the second label of the second training sample; determining the first loss value corresponding to the first training sample based on the weight value corresponding to the first training sample, the predicted value corresponding to the first training sample and the true label corresponding to the first training sample; determining the second loss value corresponding to the second training sample based on the weight value corresponding to the second training sample, the predicted value corresponding to the second training sample and the true label corresponding to the second training sample; determining the target loss value based on the first loss value and the second loss value.

[0009] Furthermore, determining the weight value corresponding to the first training sample based on the loan amount corresponding to the first training sample includes: obtaining the maximum value and the minimum value of the loan amounts corresponding to the multiple training samples; scaling the loan amount corresponding to the first training sample based on the maximum value and the minimum value to obtain the scaled value; and determining the weight value corresponding to the first training sample based on the scaled value.

[0010] Furthermore, determining the second weight value corresponding to the second training sample based on the second label of the second training sample includes: if the second label of the second training sample is a third preset value, determining the target value as the second weight value corresponding to the second training sample; if the second label of the second training sample is a fourth preset value, determining the second weight value based on the total number of the second training samples and the number of second training samples whose second label is the fourth preset value.

[0011] Furthermore, the initial risk assessment model is trained according to the target loss value to obtain the target risk assessment model, including: training the initial risk assessment model according to the target loss value to obtain a trained initial risk assessment model; evaluating the trained initial risk assessment model according to a target test set to obtain an evaluation result; if the evaluation result indicates that the evaluation is passed, the trained initial risk assessment model is determined as the target risk assessment model.

[0012] In order to achieve the above-mentioned purpose, according to another aspect of the present application, a device for determining credit risk is provided. The device comprises: an evaluation unit, which is used to perform credit risk evaluation on the target data information of the target object to be evaluated through a target risk evaluation model to obtain a risk prediction value, wherein the target data information at least includes the loan information and credit information of the target object; a first judgment unit, which is used to judge whether the target object has a first type of repayment default risk based on the risk prediction value to obtain a first judgment result; a second judgment unit, which is used to judge whether the target object has a second type of repayment default risk based on the risk prediction value if the first judgment result indicates that the target object does not have the first type of repayment default risk, to obtain a second judgment result, wherein the default time of the first type of repayment default risk and the second type of repayment default risk are different; a determination unit, which is used to determine the credit risk information corresponding to the target object based on the first judgment result and / or the second judgment result.

[0013] Furthermore, the target risk assessment model is obtained by using the following device: an acquisition unit, used to acquire a training sample set, wherein the training sample set includes multiple training samples and a true label corresponding to each training sample, each training sample is composed of historical loan information and historical credit information of the sample object, and the true label includes a first label and a second label, the first label is used to characterize whether the training sample has a first type of repayment default risk, and the second label is used to characterize whether the training sample has a second type of repayment default risk; a training unit, used to train the initial risk assessment model based on the training sample set to obtain the target risk assessment model.

[0014] Furthermore, the training unit includes: a processing subunit, used to process the training sample set through the initial risk assessment model to obtain a prediction value corresponding to each training sample; a determination subunit, used to determine a target loss value based on the prediction value and the true label; and a training subunit, used to train the initial risk assessment model based on the target loss value to obtain the target risk assessment model.

[0015] Furthermore, the determination subunit includes: a first determination module, which is used to determine the weight value corresponding to the first training sample according to the loan amount corresponding to the first training sample if the first label of the first training sample is a first preset value; a second determination module, which is used to determine the second weight value corresponding to the second training sample according to the second label of the second training sample if the first label of the second training sample is a second preset value; a third determination module, which is used to determine the first loss value corresponding to the first training sample according to the weight value corresponding to the first training sample, the predicted value corresponding to the first training sample and the true label corresponding to the first training sample; a fourth determination module, which is used to determine the second loss value corresponding to the second training sample according to the weight value corresponding to the second training sample, the predicted value corresponding to the second training sample and the true label corresponding to the second training sample; and a fifth determination module, which is used to determine the target loss value according to the first loss value and the second loss value.

[0016] Furthermore, the first determination module includes: an acquisition submodule, used to obtain the maximum and minimum values ​​of the loan amounts corresponding to the multiple training samples; a processing submodule, used to scale the loan amount corresponding to the first training sample according to the maximum and minimum values ​​to obtain the scaled value; and a first determination submodule, used to determine the weight value corresponding to the first training sample according to the scaled value.

[0017] Furthermore, the second determination module includes: a second determination submodule, which is used to determine the target value as the second weight value corresponding to the second training sample if the second label of the second training sample is a third preset value; and a third determination submodule, which is used to determine the second weight value based on the total number of the second training samples and the number of second training samples whose second labels are the fourth preset value if the second label of the second training sample is a fourth preset value.

[0018] Furthermore, the training subunit includes: a training module, used to train the initial risk assessment model according to the target loss value to obtain a trained initial risk assessment model; an evaluation module, used to evaluate the trained initial risk assessment model according to a target test set to obtain an evaluation result; and a sixth determination module, used to determine the trained initial risk assessment model as the target risk assessment model if the evaluation result indicates that the evaluation is passed.

[0019] In order to achieve the above-mentioned purpose, according to one aspect of the present application, a computer-readable storage medium is provided, wherein the storage medium stores a program, wherein when the program is executed, the device where the storage medium is located is controlled to execute any one of the above-mentioned credit risk determination methods.

[0020] In order to achieve the above objective, according to another aspect of the present application, an electronic device is provided, the electronic device comprising one or more processors and a memory, the memory being used to store the method for determining credit risk described in any one of the above implementations by the one or more processors.

[0021] Through this application, the following steps are adopted: a target risk assessment model is used to conduct a credit risk assessment on the target data information of the target object to be assessed, and a risk prediction value is obtained, wherein the target data information at least includes the loan information and credit information of the target object; based on the risk prediction value, it is judged whether the target object has a first type of repayment default risk, and a first judgment result is obtained; if the first judgment result indicates that the target object does not have the first type of repayment default risk, then based on the risk prediction value, it is judged whether the target object has a second type of repayment default risk, and a second judgment result is obtained, wherein the default time of the first type of repayment default risk and the second type of repayment default risk are different; based on the first judgment result and / or the second judgment result, the credit risk information corresponding to the target object is determined, which solves the problem that the accuracy of credit risk is relatively low due to the fixed risk inspection rule in the related art to assess the credit risk of the borrower. In this scheme, a target risk assessment model is used to conduct a credit risk assessment on the loan information and credit information of the target object to be assessed, and a risk prediction value is obtained, and then the credit risk information corresponding to the target object is determined according to the risk prediction value, and the target risk assessment model can accurately mine the characteristic information affecting credit in the loan information and credit information, thereby achieving the effect of improving the accuracy of credit risk. BRIEF DESCRIPTION OF THE DRAWINGS

[0022] The drawings constituting a part of the present application are used to provide a further understanding of the present application. The illustrative embodiments and descriptions of the present application are used to explain the present application and do not constitute an improper limitation on the present application. In the drawings:

[0023] Figure 1The process of the credit risk determination method provided in the embodiment of the present application is as follows Figure 1 ;

[0024] Figure 2 The process of the credit risk determination method provided in the embodiment of the present application is as follows Figure 2 ;

[0025] Figure 3 The process of the credit risk determination method provided in the embodiment of the present application is as follows Figure 3 ;

[0026] Figure 4 is a schematic diagram of a device for determining credit risk provided in an embodiment of the present application;

[0027] Figure 5 is a schematic diagram of an electronic device provided according to an embodiment of the present application. DETAILED DESCRIPTION

[0028] It should be noted that, in the absence of conflict, the embodiments and features in the embodiments of the present application can be combined with each other. The present application will be described in detail below with reference to the accompanying drawings and in combination with the embodiments.

[0029] In order to enable those skilled in the art to better understand the solution of the present application, the technical solution in the embodiments of the present application will be clearly and completely described below in conjunction with the drawings in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without creative work should fall within the scope of protection of the present application.

[0030] It should be noted that the terms "first", "second", etc. in the specification and claims of the present application and the above-mentioned drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequential order. It should be understood that the data used in this way can be interchanged where appropriate, so that the embodiments of the present application described here. In addition, the terms "including" and "having" and any of their variations are intended to cover non-exclusive inclusions, for example, a process, method, system, product or device comprising a series of steps or units is not necessarily limited to those steps or units clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or devices.

[0031] It should be noted that the relevant information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data for display, data for analysis, etc.) involved in this disclosure are all information and data authorized by the user or fully authorized by all parties. For example, an interface is set between this system and the relevant user or organization. Before obtaining relevant information, it is necessary to send an acquisition request to the aforementioned user or organization through the interface, and obtain relevant information after receiving the consent information fed back by the aforementioned user or organization.

[0032] The present invention is described below in conjunction with preferred implementation steps. Figure 1 The process of the credit risk determination method provided in the embodiment of the present application is as follows Figure 1 ,like Figure 1 As shown, the method comprises the following steps:

[0033] Step S101, performing credit risk assessment on target data information of a target object to be assessed by using a target risk assessment model to obtain a risk prediction value, wherein the target data information at least includes the loan information and credit information of the target object.

[0034] Optionally, the target data information of the target object to be evaluated, such as loan information and credit information, is obtained through the database of the target institution. After obtaining the above-mentioned target data information, the target data information is input into the target risk assessment model, and the target data information is subjected to credit risk assessment through the target risk assessment model to obtain a risk prediction value. In an optional embodiment, the target risk assessment model can be a deep learning model, and the risk assessment through the target risk assessment model includes the following steps: extracting features from the target data information through the target risk assessment model to obtain feature information, and then evaluating the credit risk through the feature information to obtain a corresponding risk prediction value.

[0035] Step S102, judging whether the target object has a first type of repayment default risk based on the risk prediction value, and obtaining a first judgment result;

[0036] Optionally, after obtaining the above risk prediction value, it is determined whether the user (i.e., the above target object) has the first type of repayment default risk according to the size of the risk prediction value. It should be noted that the first type of repayment default risk is a short-term default risk. For example, if the default behavior occurs in the 1st to 3rd period and the number of default days is 30, it is defined as a short-term default risk.

[0037] In an optional embodiment, if the risk prediction value is higher than 0.6, it is determined that there is a first type of repayment default risk, and if the risk prediction value is lower than 0.6, it is considered that there is no first type of repayment default risk. It should be noted that if there is a short-term default risk, there must be a medium- and long-term default risk (i.e., the second type of repayment default risk). For example, if the default occurs in the 4-N period and the number of default days is 30, it is defined as a medium- and long-term default risk.

[0038] Step S103, if the first judgment result indicates that the target object does not have the first type of repayment default risk, then judging whether the target object has the second type of repayment default risk based on the risk prediction value, and obtaining a second judgment result, wherein the first type of repayment default risk and the second type of repayment default risk have different default times;

[0039] Optionally, if the first judgment result indicates that the target object does not have the first type of repayment default risk, it is necessary to further judge whether the target object has the second type of repayment default risk to obtain a second judgment result. For example, if the risk prediction value is higher than 0.8, it is determined that the second type of repayment default risk exists, and if the risk prediction value is lower than 0.8, it is considered that the second type of repayment default risk does not exist.

[0040] For example, if the risk is low in the early stage but high in the late stage in credit risk management, it is necessary to assess whether there is a second type of repayment default risk, that is, whether there is a medium- and long-term default risk. By considering both short-term default risk and medium- and long-term default risk at the same time, it is possible to more reasonably judge the credit risk of the target object.

[0041] Step S104: determining the credit risk information corresponding to the target object according to the first judgment result and / or the second judgment result.

[0042] Optionally, based on the first judgment result and / or the second judgment result, the credit risk information corresponding to the target object is determined. For example, the credit risk information corresponding to the target object is that there is no first type of repayment default risk and no second type of repayment default risk, or there is a first type of repayment default risk, or there is no first type of repayment default risk and there is a second type of repayment default risk.

[0043] To summarize, the target risk assessment model is used to conduct credit risk assessment on the lending information and credit information of the target object to be assessed, and a risk prediction value is obtained. Then, the credit risk information corresponding to the target object is determined based on the risk prediction value. The target risk assessment model can accurately mine the characteristic information that affects credit in the lending information and credit information, thereby achieving the effect of improving the accuracy of credit risk.

[0044] Optionally, in the credit risk determination method provided in the embodiment of the present application, the following steps are adopted to obtain the target risk assessment model: obtain a training sample set, wherein the training sample set includes multiple training samples and a true label corresponding to each training sample, each training sample is composed of historical loan information and historical credit information of the sample object, and the true label includes a first label and a second label, the first label is used to characterize whether the training sample has a first type of repayment default risk, and the second label is used to characterize whether the training sample has a second type of repayment default risk; train the initial risk assessment model based on the training sample set to obtain the target risk assessment model.

[0045] In an optional embodiment, the target risk assessment model is obtained by the following steps: collecting the sample of applicants required for modeling and analysis, obtaining the application information, credit data and authorized third-party data authorized by the customer, that is, the historical loan information and historical credit information of the sample objects mentioned above. In order to improve the accuracy of the data, the obtained data can be preprocessed, including outlier processing and missing value processing.

[0046] Then, labels are added to the training samples. Customers who default in the short term are default samples, which are defined as 1; customers who do not default in the short term are normal samples, which are defined as 0. At the same time, the samples are extended and labeled. Customers who have medium- and long-term defaults are marked as '01', and customers who have no medium- and long-term defaults are marked as '00'. That is, the above-mentioned true labels include the first label and the second label. The first label is used to characterize whether the training sample has the first type of repayment default risk, and the second label is used to characterize whether the training sample has the second type of repayment default risk.

[0047] After obtaining the above training sample set, select a suitable machine learning algorithm or statistical model to build an initial risk assessment model.

[0048] Finally, the initial risk assessment model is trained using multiple training samples and the true label corresponding to each training sample to obtain the target risk assessment model.

[0049] Optionally, in the credit risk determination method provided in the embodiment of the present application, the initial risk assessment model is trained based on the training sample set to obtain the target risk assessment model, including: processing the training sample set through the initial risk assessment model to obtain the predicted value corresponding to each training sample; determining the target loss value based on the predicted value and the true label; and training the initial risk assessment model based on the target loss value to obtain the target risk assessment model.

[0050] In an optional embodiment, training the initial risk assessment model through a training sample set includes the following steps: inputting the training sample set into the initial risk assessment model, processing the training sample set through the initial risk assessment model to obtain the predicted value corresponding to each training sample, and then determining the target loss value based on the predicted value and the true label, and finally training the initial risk assessment model based on the target loss value to obtain the target risk assessment model. It should be noted that during the training process, the parameters of the model are optimized so that the loss function is minimized. After the training is completed, the model is evaluated using the test set, including calculating indicators such as accuracy, recall rate, F1 value, and generating evaluation results such as confusion matrix.

[0051] Optionally, in the method for determining credit risk provided in the embodiment of the present application, determining the target loss value based on the predicted value and the true label includes: if the first label of the first training sample is a first preset value, then determining the weight value corresponding to the first training sample based on the loan amount corresponding to the first training sample; if the first label of the second training sample is a second preset value, then determining the second weight value corresponding to the second training sample based on the second label of the second training sample; determining the first loss value corresponding to the first training sample based on the weight value corresponding to the first training sample, the predicted value corresponding to the first training sample and the true label corresponding to the first training sample; determining the second loss value corresponding to the second training sample based on the weight value corresponding to the second training sample, the predicted value corresponding to the second training sample and the true label corresponding to the second training sample; determining the target loss value based on the first loss value and the second loss value.

[0052] In an optional embodiment, the following steps are used to determine the target loss value: In order to enable the model to learn the importance of short-term default risk and medium- and long-term default risk, the corresponding weight values ​​are set. If the Label tag is equal to 1 (that is, the first label of the first training sample mentioned above is the first preset value), the borrower has a short-term default risk. At this time, it is no longer concerned about whether the borrower has a medium- and long-term default risk, because from the perspective of risk control, the borrower's risk has been controlled from the short term. At this time, more attention will be paid to the impact of high defaults. Therefore, based on the loan amount corresponding to the first training sample, the loan amount factor, that is, the above-mentioned (weight value corresponding to the first training sample), is determined.

[0053] If the Label is equal to 0 (i.e., the first label of the second training sample is the second preset value), the borrower has no short-term default risk. At this time, special attention should be paid to whether the borrower has medium- and long-term default risk to prevent missed judgment. In order to add the medium- and long-term default risk to the loss function, the medium- and long-term default factor (i.e., the second weight value mentioned above) is determined according to the second label of the second training sample.

[0054] For example, the medium- and long-term default factor and the loan amount factor are recorded as w0 and w1 respectively. Then the loss function is:

[0055] L=w1*y true *log(y pred )+w0*(1-y true )*log(1-y pred )

[0056] Among them, y true is the first label mentioned above, y pred The predicted value output by the model.

[0057] The loan amount factor w1 acts on Label=1(y true =1), the medium- and long-term default factor w0 acts on Label = 0 (y true =0). Therefore, the first loss value corresponding to the first training sample is obtained according to the weight value corresponding to the first training sample, the predicted value corresponding to the first training sample and the true label corresponding to the first training sample; and the second loss value corresponding to the second training sample is obtained according to the weight value corresponding to the second training sample, the predicted value corresponding to the second training sample and the true label corresponding to the second training sample, thereby obtaining the final target loss value.

[0058] By adding the loan amount factor to the loss function in the model, when training the model, the penalty of high loan amount is increased for the determined short-term risk samples (Label = 1), that is, the loan amount factor is used as the weight of loss to emphasize the control of high amount default risk. By adding the medium- and long-term risk factor to the loss function in the model, when training the model, the penalty of medium- and long-term risk factor is increased for the determined no short-term risk samples (Label = 0), that is, the medium- and long-term risk factor is used as the weight of loss to emphasize the control of the risk of no short-term default but medium- and long-term default. The above settings ensure that the model can learn the characteristics of medium- and long-term risks, and improve the accuracy of the subsequent determination of medium- and long-term default risks based on risk prediction values.

[0059] Optionally, in the method for determining credit risk provided in an embodiment of the present application, determining the weight value corresponding to the first training sample based on the loan amount corresponding to the first training sample includes: obtaining the maximum and minimum values ​​of the loan amounts corresponding to multiple training samples; scaling the loan amount corresponding to the first training sample based on the maximum and minimum values ​​to obtain the scaled value; and determining the weight value corresponding to the first training sample based on the scaled value.

[0060] In an optional embodiment, the following steps are used to determine the weight value corresponding to the first training sample: determine the maximum and minimum values ​​among the loan amounts corresponding to multiple training samples, and then scale the loan amount corresponding to the first training sample according to the maximum and minimum values ​​to obtain the scaled value, and then obtain the weight value corresponding to the above-mentioned first training sample.

[0061] In an alternative embodiment, Among them, LoanAmt is the loan amount. The loan amount is scaled to [1,2], which means that if the borrower has short-term risks, if the loan amount is low, the weight is close to 1, and if the loan amount is high, the weight is close to 2, and the high and low values ​​are used to emphasize the risk loss under different amounts.

[0062] Optionally, in the credit risk determination method provided in the embodiment of the present application, determining the second weight value corresponding to the second training sample based on the second label of the second training sample includes: if the second label of the second training sample is a third preset value, determining the target value as the second weight value corresponding to the second training sample; if the second label of the second training sample is a fourth preset value, determining the second weight value based on the total number of second training samples and the number of second training samples whose second label is the fourth preset value.

[0063] In an optional embodiment, the second weight value is determined by the following steps: if the second label is a third preset value (e.g., '00'), the target value (e.g., 1) is directly determined as the second weight value corresponding to the second training sample. If the second label is a fourth preset value (e.g., '01'), the second weight value is determined according to the total number of second training samples and the number of second training samples whose second label is the fourth preset value.

[0064] For example, Among them, sum('01') is the number of second training samples whose second labels are the fourth preset value, and sum('00') is the number of second training samples whose second labels are the third preset value. That is, if the borrower has neither short-term risk nor medium- and long-term risk, the weight is equal to 1. If the borrower has no short-term risk but has medium- and long-term risk, the weight is greater than 1, and the difference in value emphasizes whether there is loss due to medium- and long-term risk.

[0065] Optionally, in the credit risk determination method provided in an embodiment of the present application, the initial risk assessment model is trained according to the target loss value to obtain the target risk assessment model, which includes: training the initial risk assessment model according to the target loss value to obtain a trained initial risk assessment model; evaluating the trained initial risk assessment model according to a target test set to obtain an evaluation result; if the evaluation result indicates that the evaluation is passed, the trained initial risk assessment model is determined as the target risk assessment model.

[0066] In an optional embodiment, when the initial risk assessment model is trained according to the target loss value, in order to improve the performance of the target risk assessment model, after obtaining the trained initial risk assessment model, the trained initial risk assessment model is evaluated through the target test set, for example, including calculating indicators such as accuracy, recall rate, F1 value, and generating confusion matrix and other evaluation results. If the evaluation result indicates that the evaluation is passed, the trained initial risk assessment model is determined as the target risk assessment model. If the evaluation fails, the model can continue to be trained, or the hyperparameters can be configured again.

[0067] In an alternative embodiment, it is possible to Figure 2 The flowchart shown implements credit risk monitoring, which specifically includes: 1. Data preparation; 2. Model design, including: Label weight design and loss function setting; 3. Training configuration, including hyperparameter configuration and resource configuration. It should be noted that the hyperparameter configuration needs to be set according to the selected algorithm. After determining the model algorithm, the optimizer, learning rate, regularization term, etc. need to be configured; resource configuration needs to be selected according to the number of samples, such as CPU / GPU, single machine multi-card, multi-machine multi-card, etc.; 4. Training process, during the training process, the parameters of the model are optimized to minimize the loss function. After the training is completed, the model is evaluated using the test set, including calculating indicators such as accuracy, recall rate, F1 value, and generating evaluation results such as confusion matrix. 5. Model application and monitoring Apply the trained model to actual business to conduct risk assessment and decision support for new borrowers. At the same time, the model needs to be regularly tested and updated to adapt to changes in market and customer risks.

[0068] In an alternative embodiment, it is possible to Figure 3The flowchart shown implements credit risk monitoring, S301, performing credit risk assessment on the target data information of the target object to be assessed through a weighted target risk assessment model (i.e., the above-mentioned target risk assessment model) to obtain a risk prediction probability value; S302, judging whether the target object has a first type of repayment default risk based on the risk prediction probability value, and obtaining a first judgment result after weighted optimization of the first weight (i.e., the weight corresponding to the first type of repayment default risk); S303, if the first judgment result indicates that the target object does not have a first type of repayment default risk, judging whether the target object has a second type of repayment default risk based on the risk prediction probability value, and obtaining a second judgment result after weighted optimization of the second weight (i.e., the weight corresponding to the second type of repayment default risk); S304, determining the credit risk label corresponding to the target object based on the first judgment result and / or the second judgment result, that is, the credit risk information corresponding to the target object is that there is no first type of repayment default risk and no second type of repayment default risk, or there is a first type of repayment default risk, or there is no first type of repayment default risk and there is a second type of repayment default risk.

[0069] The method for determining credit risk provided by the embodiment of the present application is to conduct credit risk assessment on the target data information of the target object to be assessed through the target risk assessment model to obtain a risk prediction value, wherein the target data information at least includes the loan information and credit information of the target object; according to the risk prediction value, it is judged whether the target object has a first type of repayment default risk, and a first judgment result is obtained; if the first judgment result indicates that the target object does not have the first type of repayment default risk, then it is judged whether the target object has a second type of repayment default risk according to the risk prediction value, and a second judgment result is obtained, wherein the default time of the first type of repayment default risk and the second type of repayment default risk are different; according to the first judgment result and / or the second judgment result, the credit risk information corresponding to the target object is determined, which solves the problem that the accuracy of credit risk is relatively low due to the fixed risk inspection rule in the related art to assess the credit risk of the borrower. In this scheme, the credit risk assessment is conducted on the loan information and credit information of the target object to be assessed through the target risk assessment model to obtain a risk prediction value, and then the credit risk information corresponding to the target object is determined according to the risk prediction value, and the target risk assessment model can accurately mine the characteristic information affecting credit in the loan information and credit information, thereby achieving the effect of improving the accuracy of credit risk.

[0070] It should be noted that the steps shown in the flowcharts of the accompanying drawings can be executed in a computer system such as a set of computer executable instructions, and that, although a logical order is shown in the flowcharts, in some cases, the steps shown or described can be executed in an order different from that shown here.

[0071] The present application also provides a device for determining credit risk. It should be noted that the device for determining credit risk in the present application can be used to execute the method for determining credit risk provided in the present application. The following is an introduction to the device for determining credit risk provided in the present application.

[0072] Figure 4 is a schematic diagram of a device for determining credit risk according to an embodiment of the present application. Figure 4 As shown, the device includes: an evaluation unit 401, a first judgment unit 402, a second judgment unit 403 and a determination unit 404.

[0073] An evaluation unit 401 is used to perform a credit risk evaluation on target data information of a target object to be evaluated by using a target risk evaluation model to obtain a risk prediction value, wherein the target data information at least includes the loan information and credit information of the target object;

[0074] The first judgment unit 402 is used to judge whether the target object has a first type of repayment default risk according to the risk prediction value, and obtain a first judgment result;

[0075] The second judgment unit 403 is used to judge whether the target object has a second type of repayment default risk according to the risk prediction value if the first judgment result indicates that the target object does not have a first type of repayment default risk, and obtain a second judgment result, wherein the first type of repayment default risk and the second type of repayment default risk have different default times;

[0076] The determination unit 404 is used to determine the credit risk information corresponding to the target object according to the first judgment result and / or the second judgment result.

[0077] The credit risk determination device provided in the embodiment of the present application performs credit risk assessment on the target data information of the target object to be assessed through the assessment unit 401 through the target risk assessment model to obtain a risk prediction value, wherein the target data information at least includes the loan information and credit information of the target object; the first judgment unit 402 is used to judge whether the target object has a first type of repayment default risk based on the risk prediction value to obtain a first judgment result; the second judgment unit 403 judges whether the target object has a second type of repayment default risk based on the risk prediction value if the first judgment result indicates that the target object does not have the first type of repayment default risk, and obtains a second judgment result, wherein the default time of the first type of repayment default risk and the second type of repayment default risk are different; the determination unit 404 determines the credit risk information corresponding to the target object based on the first judgment result and / or the second judgment result, thereby solving the problem of low accuracy of credit risk caused by assessing the credit risk of the borrower through fixed risk inspection rules in the related art. In this scheme, a target risk assessment model is used to conduct a credit risk assessment on the loan information and credit information of the target object to be assessed, and a risk prediction value is obtained. Then, the credit risk information corresponding to the target object is determined based on the risk prediction value. The target risk assessment model can accurately mine the characteristic information that affects credit in the loan information and credit information, thereby achieving the effect of improving the accuracy of credit risk.

[0078] Optionally, in the credit risk determination device provided in the embodiment of the present application, the target risk assessment model is obtained by using the following device: an acquisition unit, used to acquire a training sample set, wherein the training sample set includes multiple training samples and a true label corresponding to each training sample, each training sample is composed of historical loan information and historical credit information of the sample object, and the true label includes a first label and a second label, the first label is used to characterize whether the training sample has a first type of repayment default risk, and the second label is used to characterize whether the training sample has a second type of repayment default risk; a training unit, used to train the initial risk assessment model based on the training sample set to obtain the target risk assessment model.

[0079] Optionally, in the credit risk determination device provided in the embodiment of the present application, the training unit includes: a processing subunit, used to process the training sample set through the initial risk assessment model to obtain a predicted value corresponding to each training sample; a determination subunit, used to determine the target loss value based on the predicted value and the true label; a training subunit, used to train the initial risk assessment model based on the target loss value to obtain a target risk assessment model.

[0080] Optionally, in the credit risk determination device provided in the embodiment of the present application, the determination subunit includes: a first determination module, which is used to determine the weight value corresponding to the first training sample according to the loan amount corresponding to the first training sample if the first label of the first training sample is a first preset value; a second determination module, which is used to determine the second weight value corresponding to the second training sample according to the second label of the second training sample if the first label of the second training sample is a second preset value; a third determination module, which is used to determine the first loss value corresponding to the first training sample according to the weight value corresponding to the first training sample, the predicted value corresponding to the first training sample and the true label corresponding to the first training sample; a fourth determination module, which is used to determine the second loss value corresponding to the second training sample according to the weight value corresponding to the second training sample, the predicted value corresponding to the second training sample and the true label corresponding to the second training sample; and a fifth determination module, which is used to determine the target loss value based on the first loss value and the second loss value.

[0081] Optionally, in the credit risk determination device provided in the embodiment of the present application, the first determination module includes: an acquisition submodule, used to obtain the maximum and minimum values ​​of the loan amounts corresponding to multiple training samples; a processing submodule, used to scale the loan amount corresponding to the first training sample according to the maximum and minimum values ​​to obtain the scaled value; the first determination submodule, used to determine the weight value corresponding to the first training sample based on the scaled value.

[0082] Optionally, in the credit risk determination device provided in the embodiment of the present application, the second determination module includes: a second determination submodule, used to determine the target value as the second weight value corresponding to the second training sample if the second label of the second training sample is a third preset value; and a third determination submodule, used to determine the second weight value based on the total number of second training samples and the number of second training samples whose second labels are the fourth preset value if the second label of the second training sample is a fourth preset value.

[0083] Optionally, in the credit risk determination device provided in the embodiment of the present application, the training subunit includes: a training module, used to train the initial risk assessment model according to the target loss value to obtain the trained initial risk assessment model; an evaluation module, used to evaluate the trained initial risk assessment model according to the target test set to obtain an evaluation result; a sixth determination module, used to determine the trained initial risk assessment model as the target risk assessment model if the evaluation result indicates that the evaluation is passed.

[0084] The credit risk determination device includes a processor and a memory. The above-mentioned evaluation unit 401, first judgment unit 402, second judgment unit 403 and determination unit 404 are all stored in the memory as program units, and the processor executes the above-mentioned program units stored in the memory to realize corresponding functions.

[0085] The processor contains a kernel, which calls the corresponding program unit from the memory. One or more kernels can be set, and accurate assessment of credit risk can be achieved by adjusting kernel parameters.

[0086] The memory may include non-permanent memory in a computer-readable medium, random access memory (RAM) and / or non-volatile memory, such as read-only memory (ROM) or flash RAM, and the memory includes at least one memory chip.

[0087] An embodiment of the present invention provides a computer-readable storage medium having a program stored thereon, and the program implements a method for determining credit risk when executed by a processor.

[0088] An embodiment of the present invention provides a processor, which is used to run a program, wherein the program executes a method for determining credit risk when running.

[0089] like Figure 5 As shown, an embodiment of the present invention provides an electronic device, the device includes a processor, a memory, and a program stored in the memory and executable on the processor, and the processor implements the following steps when executing the program: performing credit risk assessment on target data information of a target object to be assessed through a target risk assessment model to obtain a risk prediction value, wherein the target data information at least includes the target object's loan information and credit information; judging whether the target object has a first type of repayment default risk based on the risk prediction value to obtain a first judgment result; if the first judgment result indicates that the target object does not have a first type of repayment default risk, judging whether the target object has a second type of repayment default risk based on the risk prediction value to obtain a second judgment result, wherein the first type of repayment default risk and the second type of repayment default risk have different default times; determining the credit risk information corresponding to the target object based on the first judgment result and / or the second judgment result.

[0090] Optionally, the target risk assessment model is obtained by the following steps: obtaining a training sample set, wherein the training sample set includes multiple training samples and a true label corresponding to each training sample, each training sample is composed of historical loan information and historical credit information of the sample object, and the true label includes a first label and a second label, the first label is used to characterize whether the training sample has a first type of repayment default risk, and the second label is used to characterize whether the training sample has a second type of repayment default risk; training the initial risk assessment model based on the training sample set to obtain the target risk assessment model.

[0091] Optionally, training the initial risk assessment model based on the training sample set to obtain the target risk assessment model includes: processing the training sample set through the initial risk assessment model to obtain a predicted value corresponding to each training sample; determining a target loss value based on the predicted value and the true label; and training the initial risk assessment model based on the target loss value to obtain the target risk assessment model.

[0092] Optionally, determining the target loss value based on the predicted value and the true label includes: if the first label of the first training sample is a first preset value, determining the weight value corresponding to the first training sample based on the loan amount corresponding to the first training sample; if the first label of the second training sample is a second preset value, determining the second weight value corresponding to the second training sample based on the second label of the second training sample; determining the first loss value corresponding to the first training sample based on the weight value corresponding to the first training sample, the predicted value corresponding to the first training sample and the true label corresponding to the first training sample; determining the second loss value corresponding to the second training sample based on the weight value corresponding to the second training sample, the predicted value corresponding to the second training sample and the true label corresponding to the second training sample; determining the target loss value based on the first loss value and the second loss value.

[0093] Optionally, determining the weight value corresponding to the first training sample based on the loan amount corresponding to the first training sample includes: obtaining the maximum and minimum values ​​of the loan amounts corresponding to multiple training samples; scaling the loan amount corresponding to the first training sample based on the maximum and minimum values ​​to obtain a scaled value; and determining the weight value corresponding to the first training sample based on the scaled value.

[0094] Optionally, determining the second weight value corresponding to the second training sample based on the second label of the second training sample includes: if the second label of the second training sample is a third preset value, determining the target value as the second weight value corresponding to the second training sample; if the second label of the second training sample is a fourth preset value, determining the second weight value based on the total number of second training samples and the number of second training samples whose second label is the fourth preset value.

[0095] Optionally, the initial risk assessment model is trained according to the target loss value to obtain the target risk assessment model, including: training the initial risk assessment model according to the target loss value to obtain a trained initial risk assessment model; evaluating the trained initial risk assessment model according to the target test set to obtain an evaluation result; if the evaluation result indicates that the evaluation is passed, the trained initial risk assessment model is determined as the target risk assessment model.

[0096] The devices in this article can be servers, PCs, PADs, mobile phones, etc.

[0097] The present application also provides a computer program product, which, when executed on a data processing device, is suitable for executing a program that is initialized with the following method steps: performing a credit risk assessment on the target data information of the target object to be assessed through a target risk assessment model to obtain a risk prediction value, wherein the target data information at least includes the target object's loan information and credit information; judging whether the target object has a first type of repayment default risk based on the risk prediction value to obtain a first judgment result; if the first judgment result indicates that the target object does not have the first type of repayment default risk, judging whether the target object has a second type of repayment default risk based on the risk prediction value to obtain a second judgment result, wherein the first type of repayment default risk and the second type of repayment default risk have different default times; determining the credit risk information corresponding to the target object based on the first judgment result and / or the second judgment result.

[0098] Optionally, the target risk assessment model is obtained by the following steps: obtaining a training sample set, wherein the training sample set includes multiple training samples and a true label corresponding to each training sample, each training sample is composed of historical loan information and historical credit information of the sample object, and the true label includes a first label and a second label, the first label is used to characterize whether the training sample has a first type of repayment default risk, and the second label is used to characterize whether the training sample has a second type of repayment default risk; training the initial risk assessment model based on the training sample set to obtain the target risk assessment model.

[0099] Optionally, training the initial risk assessment model based on the training sample set to obtain the target risk assessment model includes: processing the training sample set through the initial risk assessment model to obtain a predicted value corresponding to each training sample; determining a target loss value based on the predicted value and the true label; and training the initial risk assessment model based on the target loss value to obtain the target risk assessment model.

[0100] Optionally, determining the target loss value based on the predicted value and the true label includes: if the first label of the first training sample is a first preset value, determining the weight value corresponding to the first training sample based on the loan amount corresponding to the first training sample; if the first label of the second training sample is a second preset value, determining the second weight value corresponding to the second training sample based on the second label of the second training sample; determining the first loss value corresponding to the first training sample based on the weight value corresponding to the first training sample, the predicted value corresponding to the first training sample and the true label corresponding to the first training sample; determining the second loss value corresponding to the second training sample based on the weight value corresponding to the second training sample, the predicted value corresponding to the second training sample and the true label corresponding to the second training sample; determining the target loss value based on the first loss value and the second loss value.

[0101] Optionally, determining the weight value corresponding to the first training sample based on the loan amount corresponding to the first training sample includes: obtaining the maximum and minimum values ​​of the loan amounts corresponding to multiple training samples; scaling the loan amount corresponding to the first training sample based on the maximum and minimum values ​​to obtain a scaled value; and determining the weight value corresponding to the first training sample based on the scaled value.

[0102] Optionally, determining the second weight value corresponding to the second training sample based on the second label of the second training sample includes: if the second label of the second training sample is a third preset value, determining the target value as the second weight value corresponding to the second training sample; if the second label of the second training sample is a fourth preset value, determining the second weight value based on the total number of second training samples and the number of second training samples whose second label is the fourth preset value.

[0103] Optionally, the initial risk assessment model is trained according to the target loss value to obtain the target risk assessment model, including: training the initial risk assessment model according to the target loss value to obtain a trained initial risk assessment model; evaluating the trained initial risk assessment model according to the target test set to obtain an evaluation result; if the evaluation result indicates that the evaluation is passed, the trained initial risk assessment model is determined as the target risk assessment model.

[0104] Those skilled in the art will appreciate that the embodiments of the present application may be provided as methods, systems, or computer program products. Therefore, the present application may adopt the form of a complete hardware embodiment, a complete software embodiment, or an embodiment in combination with software and hardware. Moreover, the present application may adopt the form of a computer program product implemented in one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) that include computer-usable program code.

[0105] The present application is described with reference to the flowcharts and / or block diagrams of the methods, devices (systems), and computer program products according to the embodiments of the present application. It should be understood that each process and / or box in the flowchart and / or block diagram, as well as the combination of the processes and / or boxes in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to generate a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0106] These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.

[0107] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process. Figure 1 A process or multiple processes and / or boxes Figure 1 The steps for the functions specified in one or more boxes.

[0108] In a typical configuration, a computing device includes one or more processors (CPU), input / output interfaces, network interfaces, and memory.

[0109] The memory may include non-permanent memory in a computer-readable medium, random access memory (RAM) and / or non-volatile memory in the form of read-only memory (ROM) or flash RAM. The memory is an example of a computer-readable medium.

[0110] Computer readable media include permanent and non-permanent, removable and non-removable media that can be implemented by any method or technology to store information. Information can be computer readable instructions, data structures, program modules or other data. Examples of computer storage media include, but are not limited to, phase change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technology, compact disk read-only memory (CD-ROM), digital versatile disk (DVD) or other optical storage, magnetic cassettes, magnetic tape disk storage or other magnetic storage devices or any other non-transmission media that can be used to store information that can be accessed by a computing device. As defined herein, computer readable media does not include temporary computer readable media (transitory media), such as modulated data signals and carrier waves.

[0111] It should also be noted that the terms "include", "comprises" or any other variations thereof are intended to cover non-exclusive inclusion, so that a process, method, commodity 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, commodity or device. In the absence of more restrictions, the elements defined by the sentence "comprises a ..." do not exclude the existence of other identical elements in the process, method, commodity or device including the elements.

[0112] Those skilled in the art will appreciate that the embodiments of the present application may be provided as methods, systems or computer program products. Therefore, the present application may adopt the form of a complete hardware embodiment, a complete software embodiment or an embodiment in combination with software and hardware. Moreover, the present application may adopt the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) that contain computer-usable program code.

[0113] The above are only embodiments of the present application and are not intended to limit the present application. For those skilled in the art, the present application may have various changes and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present application should be included within the scope of the claims of the present application.

Claims

1. A method for determining credit risk, characterized in that: include: Performing credit risk assessment on target data information of a target object to be assessed by using a target risk assessment model to obtain a risk prediction value, wherein the target data information at least includes the loan information and credit information of the target object; According to the risk prediction value, determining whether the target object has a first type of repayment default risk, and obtaining a first determination result; If the first judgment result indicates that the target object does not have a first type of repayment default risk, then judging whether the target object has a second type of repayment default risk based on the risk prediction value to obtain a second judgment result, wherein the first type of repayment default risk and the second type of repayment default risk have different default times; Determine the credit risk information corresponding to the target object based on the first judgment result and / or the second judgment result.

2. The method according to claim 1, characterized in that The target risk assessment model is obtained by following the steps below: Acquire a training sample set, wherein the training sample set includes a plurality of training samples and a true label corresponding to each training sample, each training sample is composed of historical loan information and historical credit information of the sample object, and the true label includes a first label and a second label, the first label is used to characterize whether the training sample has a first type of repayment default risk, and the second label is used to characterize whether the training sample has a second type of repayment default risk; The initial risk assessment model is trained according to the training sample set to obtain the target risk assessment model.

3. The method according to claim 2, characterized in that The initial risk assessment model is trained according to the training sample set to obtain the target risk assessment model, including: Processing the training sample set by the initial risk assessment model to obtain a prediction value corresponding to each training sample; Determining a target loss value based on the predicted value and the true label; The initial risk assessment model is trained according to the target loss value to obtain the target risk assessment model.

4. The method according to claim 3, characterized in that Determining a target loss value according to the predicted value and the true label includes: If the first label of the first training sample is a first preset value, determining a weight value corresponding to the first training sample according to the loan amount corresponding to the first training sample; If the first label of the second training sample is a second preset value, determining a second weight value corresponding to the second training sample according to the second label of the second training sample; Determine a first loss value corresponding to the first training sample according to a weight value corresponding to the first training sample, a predicted value corresponding to the first training sample, and a true label corresponding to the first training sample; Determine a second loss value corresponding to the second training sample according to the weight value corresponding to the second training sample, the predicted value corresponding to the second training sample, and the true label corresponding to the second training sample; The target loss value is determined according to the first loss value and the second loss value.

5. The method according to claim 4, characterized in that Determining the weight value corresponding to the first training sample according to the loan amount corresponding to the first training sample includes: Obtaining the maximum and minimum values ​​of the loan amounts corresponding to the multiple training samples; Scaling the loan amount corresponding to the first training sample according to the maximum value and the minimum value to obtain a scaled value; A weight value corresponding to the first training sample is determined according to the scaled value.

6. The method according to claim 4, characterized in that Determining a second weight value corresponding to the second training sample according to the second label of the second training sample includes: If the second label of the second training sample is a third preset value, determining the target value as a second weight value corresponding to the second training sample; If the second label of the second training sample is a fourth preset value, the second weight value is determined according to the total number of the second training samples and the number of second training samples whose second label is the fourth preset value.

7. The method according to claim 4, characterized in that The initial risk assessment model is trained according to the target loss value to obtain the target risk assessment model, including: Training the initial risk assessment model according to the target loss value to obtain a trained initial risk assessment model; Evaluating the trained initial risk assessment model according to the target test set to obtain an evaluation result; If the assessment result indicates that the assessment is passed, the trained initial risk assessment model is determined as the target risk assessment model.

8. A device for determining credit risk, characterized in that: include: An evaluation unit, configured to perform a credit risk evaluation on target data information of a target object to be evaluated by using a target risk evaluation model to obtain a risk prediction value, wherein the target data information at least includes the loan information and credit information of the target object; A first judgment unit is used to judge whether the target object has a first type of repayment default risk according to the risk prediction value, and obtain a first judgment result; a second judgment unit, configured to judge whether the target object has a second type of repayment default risk according to the risk prediction value if the first judgment result indicates that the target object does not have a first type of repayment default risk, and obtain a second judgment result, wherein the first type of repayment default risk and the second type of repayment default risk have different default times; A determination unit is used to determine the credit risk information corresponding to the target object according to the first judgment result and / or the second judgment result.

9. A computer-readable storage medium, characterized in that: The computer-readable storage medium includes a stored program, wherein when the program is executed, the storage medium is controlled to execute the credit risk determination method according to any one of claims 1 to 7 on a device.

10. An electronic device, characterized in that: The invention comprises one or more processors and a memory, wherein the memory is used to store one or more programs, wherein when the one or more programs are executed by the one or more processors, the one or more processors implement the credit risk determination method according to any one of claims 1 to 7.

Images