Risk assessment method, model training method, device, equipment and storage medium

By combining the information of users who have passed and failed pre-audit, the problem of inaccurate risk assessment results in the existing technology is solved, and more accurate and generalized risk assessment is achieved, and pre-loan risk assessment is improved. efficiency and accuracy.

CN111882140BActive Publication Date: 2024-11-12MASHANG CONSUMER FINANCE CO LTD
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN201911302264.1
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2019-12-17
Publication Date
2024-11-12
Estimated Expiration
2039-12-17

AI Technical Summary

Technical Problem

During the construction process, the existing pre-loan risk model only considers user information that has passed the pre-audit, resulting in inaccurate risk assessment results.

Method used

By obtaining the first training sample set (user information passed the pre-audit) and the second training sample set (user information not passed the pre-audit) and selecting the target training sample set from the second training sample set, using the XGBoost model to train risk assessment The model, considers the information of the user through and fails to pre-audit user, increases the representativeness and generalization of the training sample.

Benefits of technology

It improves the accuracy and generalization ability of the risk assessment model, reduces the bad debt rate, and improves the accuracy of the evaluation results. Compared with the traditional method, the AUC indicator is increased by 5% and the KS indicator is increased by 8%.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN111882140B_ABST
    Figure CN111882140B_ABST
Patent Text Reader

Abstract

The present invention discloses a risk assessment method, a model training method, an apparatus, a device and a storage medium, which relate to the field of data processing technology, to solve the problem of inaccurate risk assessment results for users. The method comprises: obtaining a first training sample set and a second training sample set; wherein the samples in the first training sample set are information of users who have passed the pre-examination, and the samples in the second training sample set are information of users who have not passed the pre-examination; selecting a target training sample set from the second training sample set; and using the first training sample set and the target training sample set to perform model training to obtain a risk assessment model. The embodiment of the present invention can obtain relatively accurate risk assessment results.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to the field of data processing technology, and in particular to a risk assessment method, a model training method, a device, a equipment and a storage medium. Background Art

[0002] Risk control capability is the core of a financial company. Faced with a large number of loan applicants, high-quality users will repay on time and bring profits to the company; while low-quality users maliciously owe money, which will cause huge economic losses to the company. Only with good risk control capabilities can we accurately distinguish between high-quality and low-quality users and bring economic benefits to the company.

[0003] As the number of loan application users continues to increase, the traditional manual approval and credit granting form is gradually replaced by risk models represented by the scorecard model due to its low efficiency and high cost. By building a pre-loan risk model and automating the approval process, the approval efficiency can be effectively improved and the bad debt rate can be reduced. However, in the process of building the current pre-loan risk model, only users who have passed the pre-review are taken into consideration, so the evaluation results obtained using the risk model are inaccurate. Summary of the invention

[0004] Embodiments of the present invention provide a risk assessment method, a model training method, an apparatus, a device and a storage medium to solve the problem of inaccurate risk assessment results for users.

[0005] In a first aspect, an embodiment of the present invention provides a risk assessment model training method, comprising:

[0006] Obtaining a first training sample set and a second training sample set; wherein the samples in the first training sample set are information of users who have passed the pre-examination, and the samples in the second training sample set are information of users who have not passed the pre-examination;

[0007] Selecting a target training sample set from the second training sample set;

[0008] The first training sample set and the target training sample set are used to perform model training to obtain a risk assessment model.

[0009] In a second aspect, an embodiment of the present invention further provides a risk assessment method, comprising:

[0010] Obtain information about the user to be evaluated;

[0011] Using the information of the user to be evaluated as input of the risk evaluation model to obtain the risk evaluation result of the user to be evaluated;

[0012] Wherein, the risk assessment model is obtained according to the aforementioned risk assessment model training method.

[0013] In a third aspect, an embodiment of the present invention further provides a risk assessment model training device, comprising:

[0014] A first acquisition module is used to acquire a first training sample set and a second training sample set; wherein the samples in the first training sample set are information of users who have passed the pre-examination, and the samples in the second training sample set are information of users who have not passed the pre-examination;

[0015] A selection module, configured to select a target training sample set from the second training sample set;

[0016] The training module is used to perform model training using the first training sample set and the target training sample set to obtain a risk assessment model.

[0017] In a fourth aspect, an embodiment of the present invention further provides a risk assessment device, including:

[0018] The acquisition module is used to obtain the information of the user to be evaluated;

[0019] A processing module, used to use the information of the user to be evaluated as an input of a risk evaluation model to obtain a risk evaluation result of the user to be evaluated;

[0020] Wherein, the risk assessment model is obtained according to the aforementioned risk assessment model training method.

[0021] In a fifth aspect, an embodiment of the present invention further provides an electronic device, comprising: a memory, a processor, and a program stored in the memory and executable on the processor, wherein when the processor executes the program, the steps in the method of the first aspect or the second aspect as described above are implemented.

[0022] In a sixth aspect, an embodiment of the present invention further provides a computer-readable storage medium, on which a computer program is stored. When the computer program is executed by a processor, the steps in the method of the first aspect or the second aspect as described above are implemented.

[0023] In an embodiment of the present invention, a risk assessment model is trained using a target training sample set in a first training sample set and a second training sample set, wherein the samples in the first training sample set are information of users who have passed the pre-examination, the samples in the second training sample set are information of users who have not passed the pre-examination, and the target training sample set is selected from the second training sample set. Since the information of users who have passed the pre-examination and the information of users who have not passed the pre-examination are both considered as training samples in the training model, and further screening is performed in the training samples of the information of users who have not passed the pre-examination, the training samples include a wider range and are more representative, so that the risk assessment model obtained by training has stronger generalization, therefore, a more accurate risk assessment model can be obtained using the solution of the embodiment of the present invention, thereby obtaining a more accurate risk assessment result. BRIEF DESCRIPTION OF THE DRAWINGS

[0024] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings required for use in the description of the embodiments of the present invention will be briefly introduced below. Obviously, the accompanying drawings in the following description are only some embodiments of the present invention. For ordinary technicians in this field, other accompanying drawings can be obtained based on these accompanying drawings without paying creative labor.

[0025] Figure 1 This is one of the flow charts of the risk assessment model training method provided by an embodiment of the present invention;

[0026] Figure 2 This is the second flow chart of the risk assessment model training method provided by the embodiment of the present invention;

[0027] Figure 3 is a flow chart of a risk assessment method provided by an embodiment of the present invention;

[0028] Figure 4 is a structural diagram of a risk assessment model training device provided by an embodiment of the present invention;

[0029] Figure 5 is a structural diagram of a risk assessment device provided by an embodiment of the present invention;

[0030] Figure 6 It is one of the structural diagrams of the electronic device provided by the embodiment of the present invention;

[0031] Figure 7 This is the second structural diagram of the electronic device provided by the embodiment of the present invention. DETAILED DESCRIPTION

[0032] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0033] See also Figure 1 , Figure 1 is a flow chart of the risk assessment model training method provided by an embodiment of the present invention. Figure 1 As shown, the following steps are included:

[0034] Step 101: Obtain a first training sample set and a second training sample set.

[0035] The samples in the first training sample set are information of users who have passed the pre-examination, and the samples in the second training sample set are information of users who have not passed the pre-examination.

[0036] The pre-audit may be a loan audit. In practical applications, different users may be audited according to existing loan audit methods to determine users who have passed the audit and users who have not passed the audit. Here, the information of users who have passed the audit is used to form a first training sample set, and the information of users who have not passed the audit is used to form a second training sample set. The user information may be, for example, the user's name, age, loan record, repayment information, etc.

[0037] Furthermore, the first training sample is a first training sample with a label added; wherein the label includes a first label and a second label, the first label is used to indicate that the repayment performance meets the preset requirements, and the second label is used to indicate that the repayment performance does not meet the preset requirements. For example, the first label indicates that the user repays on time, and the second label indicates that the user has overdue repayments.

[0038] Step 102: Select a target training sample set from the second training sample set.

[0039] In an embodiment of the present invention, in order to obtain a more accurate risk assessment model, a target training sample set may also be selected from the second training sample set. When selecting the target training sample set from the second training sample set, factors such as the qualifications of each sample may be considered. The qualification score is a parameter used to indicate whether the user repays on time. The higher the qualification score, the higher the possibility that the user repays on time, otherwise the lower. In an embodiment of the present invention, the qualification score of each training sample in the target training sample set is less than or equal to the preset score. The preset score can be set arbitrarily. For example, the preset score can be set to 0. Then, here, the qualification score of each training sample in the target training sample set is 0.

[0040] Specifically, when selecting a target training sample set from the second training sample set, the qualification score of each training sample in the second training sample set is obtained, and then, training samples with qualification scores less than or equal to a preset score are selected from the second training sample set to form the target training sample set. In practical applications, the second training sample set is input into a qualification screening model to obtain the qualification scores of each training sample in the second training sample set. Specifically, the existing qualification screening model can be used to obtain the qualification scores of each training sample in the second training sample set. Alternatively, in an embodiment of the present invention, the qualification screening model can also be trained based on historical data or samples to obtain the qualification scores of each training sample in the second training sample set.

[0041] When training the qualification screening model, a third set of samples to be trained can be obtained. Among them, each sample in the third set of samples to be trained can be composed of user information arbitrarily selected from historical users. Based on the information of these historical users, information such as their repayment performance can be obtained. Then, based on the repayment performance of each sample in the third set of samples to be trained, the qualification score of each sample in the third set of samples to be trained is obtained. Specifically, the qualification score of each sample can be determined based on the correspondence between repayment performance and qualification score. For example, for users with no overdue repayment record, their qualification score can be set to 10 points; for users with overdue repayment record, if the overdue repayment record is 1, their qualification score can be set to 9 points, etc. Finally, based on the third set of samples to be trained and the qualification scores of each sample in the third set of samples to be trained, XGBoost (Extreme Gradient Boosting) is used for model training to obtain the qualification screening model.

[0042] Step 103: Perform model training using the first training sample set and the target training sample set to obtain the risk assessment model.

[0043] Specifically, when performing model training, the first training sample set and the target training sample set are input into XGBoost for model training to obtain a risk assessment model. When using the XGBoost model for training, variable screening is not required, and all variable information can be used, thereby avoiding the loss of feature information and further improving the accuracy of the obtained model.

[0044] In an embodiment of the present invention, a risk assessment model is trained using a target training sample set in a first training sample set and a second training sample set, wherein the samples in the first training sample set are information of users who have passed the pre-examination, and the samples in the second training sample set are information of users who have not passed the pre-examination, and the target training sample set is obtained by further screening in the second training sample set. Since the information of users who have passed the pre-examination and the information of users who have not passed the pre-examination are both considered as training samples in the training model, and further screening is performed in the training samples of the information of users who have not passed the pre-examination, the training samples include a wider range and are more representative. Therefore, a more accurate risk assessment model can be obtained using the solution of the embodiment of the present invention, thereby obtaining a more accurate risk assessment result.

[0045] Furthermore, the target training sample set is screened in the second training sample set based on the qualification score reaching the preset score. The further screening of the second training sample set is to screen out the sample data corresponding to a part of users with relatively good credit, and train the risk assessment model with the sample data corresponding to this part of users with relatively poor credit, so that when training the risk assessment model, the information of high-quality users and low-quality users learned is more accurate, and some users with good credit will not be easily identified as bad users. Therefore, similarly, a more accurate risk assessment model can be obtained using the solution of the embodiment of the present invention, thereby obtaining a more accurate risk assessment result.

[0046] See also Figure 2 , Figure 2 is a flow chart of the risk assessment model training method provided by an embodiment of the present invention. Figure 2 As shown, the following steps are included:

[0047] Step 201: Obtain a first training sample set and a second training sample set.

[0048] According to the approval results of users, the users who applied are divided into a user group whose applications are approved and a user group whose applications are rejected. The information of users in the user group whose applications are approved constitutes a first training sample set, and the information of users in the user group whose applications are rejected constitutes a second training sample set.

[0049] Step 202: Add labels to samples in the first training sample set.

[0050] For users who have applied for the user group, repayment label information is added to the user according to the repayment performance, which is used for subsequent model training. In this embodiment of the present invention, how to add labels and the specific form of the labels are not limited, as long as users with different repayment performance can be distinguished.

[0051] Step 203: Screen the second training sample set to obtain a target training sample set.

[0052] In order to increase the stability of the obtained model effect, in an embodiment of the present invention, the second training sample set is screened to make the labeling more reasonable. Specifically, the qualification screening model is constructed using historical samples and data. For each sample in the second training sample set, the model can be used to obtain its qualification score, which is greater than or equal to 0 and less than or equal to 1. The closer the qualification score is to 1, the better the user's qualification is, and the closer it is to 0, the worse the user's qualification is.

[0053] Here, samples with a qualification score of 0 or the lowest qualification score are used as target training samples, that is, samples with the worst qualifications are regarded as "bad samples". Although these samples have no actual repayment performance, if they have borrowed money, they are likely to be overdue. Therefore, adding a "bad sample" label to them and using it for subsequent model training can make the obtained model more accurate.

[0054] In the embodiment of the present invention, the model screening method is used to screen out the user group with the worst qualifications from the user group that failed the review, and add the "bad sample" label to the subsequent model training. By adding the "bad samples" corresponding to the user group with the worst qualifications to the training samples, the sample range can be enriched and the generalization ability of the model can be increased. At the same time, since only the user group with the worst qualifications is added, the user information learned by the risk assessment model is more accurate, so the misjudgment of rejected samples can be avoided as much as possible, making the model prediction more accurate.

[0055] There is no strict order relationship between step 202 and step 203.

[0056] Step 204: Based on the labeled first training sample set and the target training sample set, an Xgboost ensemble tree model is used for training to obtain a risk assessment model.

[0057] Among them, the risk assessment model can be used for pre-loan risk assessment.

[0058] As can be seen from the above, in the model training of the embodiment of the present invention, the information of the user group with poor qualifications and rejected applications is added, and the modeling samples are enriched, so that the distribution of the training samples can be close to the distribution of the real user group, so that the model has better generalization ability and can improve the prediction ability of the model. By comparison, when using 360,000 test samples, using AUC (Area Under Curve) and KS (Kolmogorov-Smirnov) as evaluation indicators, compared with the model established by the traditional method, the AUC index of the risk assessment model trained by the embodiment of the present invention can be improved by 5%, and the KS index can be improved by 8%.

[0059] See also Figure 3 , Figure 3 is a flow chart of a risk assessment method provided by an embodiment of the present invention. Figure 3 As shown, the following steps are included:

[0060] Step 301: Obtain information of the user to be evaluated.

[0061] The information of the user to be evaluated may include name, age, repayment record, etc.

[0062] Step 302: Use the information of the user to be evaluated as input to a risk evaluation model to obtain a risk evaluation result of the user to be evaluated.

[0063] Wherein, the risk assessment model is obtained according to the aforementioned risk assessment model training method.

[0064] In an embodiment of the present invention, a risk assessment model is trained using a first training sample set and a second training sample set to which labels are added, wherein the samples in the first training sample set are information of users who have passed the pre-examination, and the samples in the second training sample set are information of users who have not passed the pre-examination. Since both the information of users who have passed the pre-examination and the information of users who have not passed the pre-examination are used as training samples in the training model, a more accurate risk assessment model can be obtained using the solution of the embodiment of the present invention, thereby obtaining a more accurate risk assessment result.

[0065] The embodiment of the present invention also provides a risk assessment model training device. Figure 4 , Figure 4 1 is a structural diagram of a risk assessment model training device provided in an embodiment of the present invention. Since the principle of solving the problem by the risk assessment model training device is similar to the risk assessment model training method in an embodiment of the present invention, the implementation of the risk assessment model training device can refer to the implementation of the method, and the repeated parts will not be repeated.

[0066] like Figure 4 As shown, the risk assessment model training device 400 includes: a first acquisition module 401, used to obtain a first training sample set and a second training sample set; wherein, the samples in the first training sample set are information of users who have passed the pre-examination, and the samples in the second training sample set are information of users who have not passed the pre-examination; a selection module 402, used to select a target training sample set from the second training sample set; a training module 403, used to perform model training using the first training sample set and the target training sample set to obtain a risk assessment model.

[0067] Optionally, the selection module 402 includes: an acquisition submodule, used to obtain the qualification score of each training sample in the second training sample set; and a selection submodule, used to select training samples with qualification scores less than or equal to a preset score from the second training sample set to form the target training sample set.

[0068] Optionally, the acquisition submodule is used to input the second training sample set into a qualification screening model to obtain the qualification score of each training sample in the second training sample set.

[0069] Optionally, the training module 403 may also be used to train the qualification screening model. In this case, the training module 403 may include:

[0070] The first acquisition submodule is used to acquire the third sample set to be trained; the second acquisition submodule is used to determine the qualification score of each sample in the third sample set to be trained; the training subunit is used to input the third sample set to be trained and the qualification score of each sample in the third sample set to be trained into extreme gradient boosting XGBoost for model training to obtain the qualification screening model.

[0071] Optionally, the training module 403 is specifically used to input the first training sample set and the target training sample set into XGBoost for model training to obtain a risk assessment model.

[0072] The device provided in the embodiment of the present invention can execute the above method embodiment, and its implementation principle and technical effect are similar, so this embodiment will not be repeated here.

[0073] The embodiment of the present invention also provides a risk assessment device. Figure 5 , Figure 5 : is a structural diagram of a risk assessment device provided in an embodiment of the present invention. Since the principle of solving the problem by the risk assessment device is similar to the risk assessment method in the embodiment of the present invention, the implementation of the risk assessment device can refer to the implementation of the method, and the repeated parts will not be repeated.

[0074] like Figure 5 As shown, the risk assessment device 500 includes:

[0075] The acquisition module 501 is used to obtain the information of the user to be evaluated; the processing module 502 is used to use the information of the user to be evaluated as the input of the risk assessment model to obtain the risk assessment result of the user to be evaluated; wherein the risk assessment model is obtained according to the above-mentioned risk assessment model training method.

[0076] The device provided in the embodiment of the present invention can execute the above method embodiment, and its implementation principle and technical effect are similar, so this embodiment will not be repeated here.

[0077] like Figure 6 As shown, the electronic device according to the embodiment of the present invention includes: a processor 600, which is used to read the program in the memory 620 and execute the following process:

[0078] Obtaining a first training sample set and a second training sample set; wherein the samples in the first training sample set are information of users who have passed the pre-examination, and the samples in the second training sample set are information of users who have not passed the pre-examination;

[0079] Selecting a target training sample set from the second training sample set;

[0080] The first training sample set and the target training sample set are used to perform model training to obtain a risk assessment model.

[0081] The transceiver 610 is configured to receive and send data under the control of the processor 600 .

[0082] Among them, Figure 6 In the embodiment, the bus architecture may include any number of interconnected buses and bridges, specifically linking together various circuits of one or more processors represented by processor 600 and memory represented by memory 620. The bus architecture may also link together various other circuits such as peripherals, voltage regulators, and power management circuits, which are well known in the art and are therefore not further described herein. The bus interface provides an interface. The transceiver 610 may be a plurality of components, namely, a transmitter and a transceiver, providing a unit for communicating with various other devices on a transmission medium. The processor 600 is responsible for managing the bus architecture and general processing, and the memory 620 may store data used by the processor 600 when performing operations.

[0083] The processor 600 is responsible for managing the bus architecture and general processing, and the memory 620 can store data used by the processor 600 when performing operations.

[0084] The processor 600 is further configured to read the program and execute the following steps:

[0085] Obtaining a qualification score for each training sample in the second training sample set;

[0086] Select training samples whose qualification scores are less than or equal to a preset score from the second training sample set as the target training sample set.

[0087] The processor 600 is further configured to read the program and execute the following steps:

[0088] The second training sample set is input into the qualification screening model to obtain the qualification score of each training sample in the second training sample set.

[0089] The processor 600 is further configured to read the program and execute the following steps:

[0090] Training the qualification screening model includes:

[0091] Obtaining a third sample set to be trained;

[0092] Determining the qualification score of each sample in the third to-be-trained sample set;

[0093] The third sample set to be trained and the qualification score of each sample in the third sample set to be trained are input into extreme gradient boosting XGBoost for model training to obtain the qualification screening model.

[0094] The processor 600 is further configured to read the program and execute the following steps:

[0095] The first training sample set and the target training sample set are input into XGBoost for model training to obtain a risk assessment model.

[0096] like Figure 7 As shown, the electronic device according to the embodiment of the present invention includes: a processor 700, which is used to read the program in the memory 720 and execute the following process:

[0097] Obtain information about the user to be evaluated;

[0098] Using the information of the user to be evaluated as input of the risk evaluation model to obtain the risk evaluation result of the user to be evaluated;

[0099] Wherein, the risk assessment model is obtained according to the aforementioned risk assessment model training method.

[0100] The transceiver 710 is configured to receive and send data under the control of the processor 700 .

[0101] Among them, Figure 7 In the embodiment, the bus architecture may include any number of interconnected buses and bridges, specifically linking together various circuits of one or more processors represented by processor 700 and memory represented by memory 720. The bus architecture may also link together various other circuits such as peripherals, voltage regulators, and power management circuits, which are well known in the art and are therefore not further described herein. The bus interface provides an interface. The transceiver 710 may be a plurality of components, namely, a transmitter and a transceiver, providing a unit for communicating with various other devices on a transmission medium. The processor 700 is responsible for managing the bus architecture and general processing, and the memory 720 may store data used by the processor 700 when performing operations.

[0102] The processor 700 is responsible for managing the bus architecture and general processing, and the memory 720 can store data used by the processor 700 when performing operations.

[0103] The embodiment of the present invention also provides a computer-readable storage medium, on which a computer program is stored, and when the computer program is executed by a processor, the various processes of the above-mentioned risk assessment model training method or risk assessment method embodiment are implemented, and the same technical effect can be achieved. To avoid repetition, it is not repeated here. The computer-readable storage medium is, for example, a read-only memory (ROM), a random access memory (RAM), a magnetic disk or an optical disk.

[0104] It should be noted that, in this article, the terms "include", "comprises" or any other variations thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device. In the absence of further restrictions, an element defined by the sentence "comprises a ..." does not exclude the existence of other identical elements in the process, method, article or device including the element.

[0105] Through the description of the above implementation methods, those skilled in the art can clearly understand that the above-mentioned embodiment methods can be implemented by means of software plus a necessary general hardware platform, and of course by hardware, but in many cases the former is a better implementation method. Based on such an understanding, the technical solution of the present invention, or the part that contributes to the prior art, can be embodied in the form of a software product, which is stored in a storage medium (such as ROM / RAM, a magnetic disk, or an optical disk), and includes a number of instructions for enabling a terminal (which can be a mobile phone, a computer, a server, an air conditioner, or a network device, etc.) to execute the methods described in each embodiment of the present invention.

[0106] The embodiments of the present invention are described above in conjunction with the accompanying drawings, but the present invention is not limited to the above-mentioned specific implementation modes, which are merely illustrative rather than restrictive. Under the guidance of the present invention, ordinary technicians in this field can also make many forms without departing from the scope of protection of the present invention and the claims, all of which are within the protection of the present invention.

Claims

1. A risk assessment model training method, characterized in that: include: Acquire a first training sample set and a second training sample set with labels; wherein the samples in the first training sample set are information of users who have passed the pre-examination, and the samples in the second training sample set are information of users who have not passed the pre-examination, and the labels include a first label and a second label, the first label is used to indicate that the repayment performance meets the preset requirements, and the second label is used to indicate that the repayment performance does not meet the preset requirements; Selecting a target training sample set from the second training sample set, and adding a label to the target training sample set, wherein the label is used to indicate that if a training sample in the target training sample set has a loan, the training sample in the target training sample set may have a repayment performance that does not meet the preset requirements; Model training is performed using the first training sample set with labels and the target training sample set with labels to obtain a risk assessment model.

2. The method according to claim 1, characterized in that The selecting a target training sample set from the second training sample set comprises: Obtaining a qualification score for each training sample in the second training sample set; Select training samples whose qualification scores are less than or equal to a preset score from the second training sample set as the target training sample set.

3. The method according to claim 2, characterized in that The obtaining the qualification score of each training sample in the second training sample set includes: The second training sample set is input into the qualification screening model to obtain the qualification score of each training sample in the second training sample set.

4. The method according to claim 3, characterized in that The method further includes: training the qualification screening model, specifically including: Obtaining a third sample set to be trained; Determining the qualification score of each sample in the third to-be-trained sample set; The third sample set to be trained and the qualification score of each sample in the third sample set to be trained are input into extreme gradient boosting XGBoost for model training to obtain the qualification screening model.

5. The method according to any one of claims 1 to 4, characterized in that: The step of performing model training using the first training sample set and the target training sample set to obtain the risk assessment model includes: The first training sample set and the target training sample set are input into XGBoost for model training to obtain a risk assessment model.

6. A risk assessment method, characterized in that: include: Obtain information about the user to be evaluated; Using the information of the user to be evaluated as input of the risk evaluation model to obtain the risk evaluation result of the user to be evaluated; Wherein, the risk assessment model is obtained according to the risk assessment model training method of any one of claims 1-5.

7. A risk assessment model training device, characterized in that: include: A first acquisition module is used to acquire a first training sample set and a second training sample set with labels; wherein the samples in the first training sample set are information of users who have passed the pre-examination, and the samples in the second training sample set are information of users who have not passed the pre-examination, and the labels include a first label and a second label, the first label is used to indicate that the repayment performance meets the preset requirements, and the second label is used to indicate that the repayment performance does not meet the preset requirements; A selection module, configured to select a target training sample set from the second training sample set, and add a label to the target training sample set, wherein the label is used to indicate that if a training sample in the target training sample set has a loan, the training sample in the target training sample set may have a repayment performance that does not meet the preset requirements; The training module is used to perform model training using the first training sample set with labels and the target training sample set with labels to obtain a risk assessment model.

8. A risk assessment device, characterized in that: include: The acquisition module is used to obtain the information of the user to be evaluated; A processing module, used to use the information of the user to be evaluated as an input of a risk evaluation model to obtain a risk evaluation result of the user to be evaluated; Wherein, the risk assessment model is obtained according to the risk assessment model training method of any one of claims 1-5.

9. An electronic device, comprising: A memory, a processor, and a program stored in the memory and executable on the processor; wherein the processor is used to read the program in the memory to implement the steps in the method as claimed in any one of claims 1 to 5; or to implement the steps in the method as claimed in claim 6.

10. A computer-readable storage medium for storing a computer program, characterized in that: When the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 5 are implemented; or the steps of the method according to claim 6 are implemented.