Risk control model construction method based on big data

Abstract

The invention discloses a risk control model construction method based on big data, and relates to the technical field of risk control, and the method comprises the steps: obtaining financial risk hidden troubles and risk reference data; obtaining a financial risk combination; obtaining a risk weight by using a combined weight analysis method; obtaining a comprehensive weight by using a weight comprehensive algorithm; obtaining a basic risk control model, training the basic risk control model by using the comprehensive weight, and obtaining a financial risk control model based on a training result; the method is used for solving the problems that in an existing risk control model construction method, multiple risks for analysis in a risk control model cannot be effectively analyzed, each risk hidden danger cannot be effectively graded based on a combination mode of the multiple risk hidden dangers, and consequently, even if a user carries out risk control, income reduction and even loss still exist.

Description

Technical Field

[0001] The present invention relates to the field of risk control technology, and specifically to a method for constructing a risk control model based on big data. Background Art

[0002] The risk control model, whose full name is risk control model, is a tool used by financial institutions in the risk management process. It uses data to build a risk management model to help financial institutions find a balance between risk and return, thereby maximizing profits. The risk control model can be applied to credit guarantee companies and financial institutions such as banks to help these institutions conduct risk control and credit management. In a broad sense, the risk control model represents any risk management model built using data. In a narrow sense, it is a mathematical model developed using algorithms such as statistics, machine learning, and even deep learning. These models help financial institutions make decisions by analyzing, predicting, and assessing risks of large amounts of data.

[0003] Existing methods for building risk control models are usually improvements in the personalization of risk control models. A general risk control model is initially built and the built risk control model is trained using a sample set to achieve the purpose of obtaining a risk control model that meets the personalized needs of users. Although this improved method can obtain a personalized risk control model, it is unable to effectively analyze multiple risks analyzed in the risk control model. When a user faces multiple risk hazards at the same time, only using a personalized general risk control model, it is impossible to effectively rate each risk hazard based on the combination of multiple risk hazards. For example, under a specific risk combination, the losses caused by some risks will increase, resulting in reduced returns or even losses even if the user performs risk control. For example, in the patent application with publication number CN112950221A, the public A method and device for establishing a risk control model and a risk control method and device have been developed. The solution is to generate a sample set corresponding to the target user by utilizing the behavioral data of the target user; the sample set is used to train the general risk control model to obtain a personalized risk control model corresponding to the target user. Other improvements in risk control model construction usually improve the control efficiency of different risks, but it is still impossible to effectively analyze the multiple risks analyzed in the risk control model. When the user faces multiple risk hazards at the same time, it is impossible to effectively rate each risk hazard based on the combination of multiple risk hazards. For example, under a specific risk combination, the losses caused by some risks will increase, resulting in reduced returns or even losses even if the user performs risk control. In view of this, it is necessary to improve the existing method for risk control model construction. Summary of the invention

[0004] The present invention aims to solve one of the technical problems in the prior art to at least a certain extent. By proposing a risk control model construction method based on big data, it is used to solve the problem that in the existing methods for building risk control models, multiple risks analyzed in the risk control model cannot be effectively analyzed. When the user faces multiple risk hazards at the same time, each risk hazard cannot be effectively rated based on the combination of multiple risk hazards. For example, under a specific risk combination method, the losses caused by some risks will increase, which will result in reduced returns or even losses even if the user performs risk control.

[0005] To achieve the above objectives, this application provides a method for building a risk control model based on big data, comprising the following steps:

[0006] Obtain multiple financial risk hazards based on big data, and obtain risk reference data based on multiple financial risk hazards; analyze the risk reference data, and obtain financial risk combinations based on the analysis results;

[0007] Use the combined weight analysis method to analyze all financial risk combinations, and obtain multiple risk weights for each financial risk potential based on the analysis results; use the weighted comprehensive algorithm based on the risk weights to obtain the comprehensive weight corresponding to each financial risk potential;

[0008] Obtain the basic risk control model based on big data, use all financial risk hazards and corresponding comprehensive weights to train the basic risk control model, and obtain the financial risk control model based on the training results.

[0009] Further, obtaining multiple financial risk hazards based on big data, and obtaining risk reference data based on multiple financial risk hazards include:

[0010] Based on big data, multiple hidden risks in the financial industry are obtained and recorded as financial hidden risks; for any financial hidden risk, multiple financial data containing financial hidden risks on the Internet are obtained based on big data and recorded as risk reference data; for any risk reference data, the financial hidden risks in the risk reference data are recorded as core hidden risks; the hidden risks other than financial hidden risks in the risk reference data are recorded as collaborative hidden risks;

[0011] Obtain the core hidden dangers and coordinated hidden dangers of all risk reference data corresponding to the financial risk hidden dangers; obtain the loss amount in all risk reference data of the financial risk hidden dangers; establish a plane rectangular coordinate system, recorded as the hidden danger analysis coordinate system, wherein the unit of the Y-axis of the hidden danger analysis coordinate system is the amount, and the coordinate points in the X-axis from the coordinate origin to the right are filled in with all the risk reference data corresponding to the financial risk hidden dangers in sequence.

[0012] Furthermore, the risk reference data is analyzed, and the financial risk portfolio is obtained based on the analysis results, including:

[0013] Based on the loss amount of each risk reference data, mark the points in the hidden danger analysis coordinate system and record them as loss points; record the number of risk reference data as n, and record (1 / n)×100% as the loss ratio;

[0014] For any two loss points, the minimum value of the two values ​​obtained by multiplying the vertical coordinates of the two loss points by the loss ratio is recorded as the loss threshold; when the difference between the vertical coordinates of the two loss points is less than the loss threshold, the horizontal coordinates of the two loss points are marked as primary comparison data and secondary comparison data; when the core hidden dangers and coordinated hidden dangers of the primary comparison data are the same as the core hidden dangers and coordinated hidden dangers of the secondary comparison data, the core hidden dangers and coordinated hidden dangers of the primary comparison data are recorded as financial risk combinations.

[0015] Furthermore, analyzing the risk reference data and obtaining the financial risk portfolio based on the analysis results also includes:

[0016] When the core hidden dangers and coordinated hidden dangers of the primary comparison data are only partially identical with the core hidden dangers and coordinated hidden dangers of the secondary comparison data, the core hidden dangers and coordinated hidden dangers of the primary comparison data and the core hidden dangers and coordinated hidden dangers of the secondary comparison data that are identical to each other shall be recorded as a financial risk combination;

[0017] When the core hidden dangers and coordinated hidden dangers of the primary comparison data are completely different from the core hidden dangers and coordinated hidden dangers of the secondary comparison data, the core hidden dangers and coordinated hidden dangers of the primary comparison data and the core hidden dangers and coordinated hidden dangers of the secondary comparison data shall be recorded as two financial risk combinations respectively.

[0018] Furthermore, analyzing the risk reference data and obtaining the financial risk portfolio based on the analysis results also includes:

[0019] For any risk reference data that is not recorded as primary comparison data or secondary comparison data, the core hidden dangers and coordinated hidden dangers corresponding to the risk reference data are recorded as risk financial combinations;

[0020] Obtain all risky financial combinations corresponding to financial risk hazards; obtain all risky financial combinations corresponding to all financial risk hazards.

[0021] Furthermore, the combined weight analysis method includes:

[0022] For any risk financial combination corresponding to all financial risk hazards, the risk weight of the core hazard in the financial risk combination is set to 1, the number of collaborative hazards in the financial risk combination is recorded as m, and the risk weight of any collaborative hazard in the financial risk combination is set to 1 / m; obtain the risk weights corresponding to the core hazards and collaborative hazards in all risk financial combinations corresponding to the financial risk hazards, where the same financial risk hazard can obtain multiple independent risk weights in multiple financial risk combinations.

[0023] Furthermore, the weighted comprehensive algorithm is used based on the risk weight to obtain the comprehensive weight corresponding to each financial risk hazard, including:

[0024] For any financial risk: the weighted comprehensive algorithm is: Among them, F is the comprehensive weight, M is the number of financial risk portfolios, and f i is the risk weight of the financial risk hazard in each financial risk portfolio, wherein, when the financial risk hazard does not obtain a risk weight in the financial risk portfolio, the risk weight of the financial risk hazard in the financial risk portfolio is set to 0;

[0025] Obtain the comprehensive weights corresponding to all financial risk hazards.

[0026] Furthermore, a basic risk control model is obtained based on big data, and all financial risk hazards and corresponding comprehensive weights are used to train the basic risk control model. The financial risk control model is obtained based on the training results, including:

[0027] A basic risk control model is obtained based on big data; all financial risk hazards are input into the basic risk control model in sequence for recording; in the basic risk control model: when the comprehensive weight of a newly recorded financial risk hazard is greater than the maximum value of the comprehensive weights of the recorded financial risk hazards, the rating of the financial risk hazard with the highest risk rating at this time is lowered, and the risk rating of the newly recorded financial risk hazard is set to the highest; when the comprehensive weight of a newly recorded financial risk hazard is less than the minimum value of the comprehensive weights of the recorded financial risk hazards, the risk rating of the financial risk hazard with the lowest risk rating at this time is increased, and the risk rating of the newly recorded financial risk hazard is set to the lowest.

[0028] Furthermore, a basic risk control model is obtained based on big data, and all financial risk hazards and corresponding comprehensive weights are used to train the basic risk control model. The financial risk control model is obtained based on the training results and further includes:

[0029] When the comprehensive weight of a newly recorded financial risk hazard is less than the maximum value of the comprehensive weights of the recorded financial risk hazards and greater than the minimum value of the comprehensive weights of the recorded financial risk hazards, the risk rating of the newly recorded financial risk hazard is set to the middle value of all risk ratings.

[0030] Furthermore, a basic risk control model is obtained based on big data, and all financial risk hazards and corresponding comprehensive weights are used to train the basic risk control model. The financial risk control model is obtained based on the training results and further includes:

[0031] When all financial risk hazards are input into the basic risk control model, the basic risk control model at this time will be recorded as the financial risk control model.

[0032] Beneficial effects of the present invention: The present invention first obtains multiple financial risk hazards and risk reference data based on big data; analyzes the risk reference data, and obtains financial risk combinations based on the analysis results; then uses the combination weight analysis method to analyze all financial risk combinations, and obtains multiple risk weights of each financial risk hazard based on the analysis results. The advantage of this is that by obtaining the financial risk combinations and risk weights, all combinations of multiple risks appearing in the same financial data can be obtained when analyzing financial risks. By obtaining the risk weights, the proportion of each financial risk hazard in the financial risk combination can be obtained, which is helpful for effectively rating the risk hazards in the financial risk hazards in subsequent analysis, thereby helping users to focus on risk hazards with higher risk ratings;

[0033] The present invention also obtains the comprehensive weight corresponding to each financial risk hazard by using a weighted comprehensive algorithm based on the risk weight; finally, a basic risk control model is obtained based on big data, and the basic risk control model is trained using all financial risk hazards and the corresponding comprehensive weights, and a financial risk control model is obtained based on the training results. The advantage of this is that by obtaining the comprehensive weight and obtaining the financial risk control model, it can be ensured that the risk hazards under different financial risk combinations can be effectively rated through analysis through the financial risk control model, and even under a specific risk combination, the losses caused by each risk hazard can still be effectively analyzed, preventing users from still having the problem of reduced returns or even losses after risk control. BRIEF DESCRIPTION OF THE DRAWINGS

[0034] Figure 1 is a flow chart of the steps of the method of the present invention;

[0035] Figure 2 A schematic diagram of a coordinate system for analyzing hidden dangers of the present invention;

[0036] Figure 3 It is a schematic structural diagram of the electronic device of the present invention. DETAILED DESCRIPTION

[0037] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0038] Example 1, first aspect, please refer to Figure 1 As shown, this application provides a method for building a risk control model based on big data, comprising the following steps:

[0039] Step S1, obtaining multiple financial risk hazards based on big data, and obtaining risk reference data based on the multiple financial risk hazards; analyzing the risk reference data, and obtaining a financial risk combination based on the analysis results;

[0040] Step S1 includes: step S101, based on big data, multiple hidden risks in the financial industry are obtained and recorded as financial hidden risks; for any financial hidden risk, multiple financial data containing financial hidden risks in the Internet are obtained based on big data and recorded as risk reference data; for any risk reference data, the financial hidden risk in the risk reference data is recorded as a core hidden risk; the hidden risk in the risk reference data other than the financial hidden risk is recorded as a collaborative hidden risk;

[0041] In the specific implementation process, for example, in the actual analysis, the financial risk hidden danger analyzed is the market exchange rate risk; then multiple financial data containing the market exchange rate risk can be obtained and recorded as risk reference data; for any risk reference data obtained: clothing export data, the financial risk hidden dangers in the clothing export data include: market exchange rate risk, market interest rate risk and product pricing risk, then the market exchange rate risk can be recorded as the core hidden danger of clothing export data, and the market interest rate risk and product pricing risk can be recorded as the coordinated hidden dangers of clothing export data;

[0042] Step S102, obtaining the core hidden dangers and coordinated hidden dangers of all risk reference data corresponding to the financial risk hidden dangers; obtaining the loss amount in all risk reference data of the financial risk hidden dangers; establishing a plane rectangular coordinate system, recorded as the hidden danger analysis coordinate system, wherein the unit of the Y axis of the hidden danger analysis coordinate system is the amount, and the coordinate points in the X axis from the coordinate origin to the right are filled in with all the risk reference data corresponding to the financial risk hidden dangers in sequence;

[0043] In the specific implementation process, for example, in a data processing, the coordinate system for hidden danger analysis is obtained as follows: Figure 2As shown, the point corresponding to the center of the circle ○ is the loss point. Through analysis, it can be obtained that the ordinates of the loss point DD1 and the loss point DD2 are both YY1, that is, the difference between the ordinates of the loss point DD1 and the loss point DD2 must be less than the loss threshold. The abscissas XX1 and XX2 of the loss point DD1 and the loss point DD2 should be recorded as the primary comparison data and the secondary comparison data for subsequent analysis;

[0044] Step S103, based on the loss amount of each risk reference data, mark the points in the hidden danger analysis coordinate system and record them as loss points; record the number of risk reference data as n, and record (1 / n)×100% as the loss ratio;

[0045] In summary of the specific implementation process, for example, in a data processing, the number of risk reference data is 4, then through calculation, it can be obtained that the loss ratio is 25%; in specific implementation, the loss ratio can be set according to the number of risk reference data that can actually be obtained, to ensure that when the number of risk reference data is large, the loss ratio is reduced, and when the number of risk reference data is small, the loss ratio is increased;

[0046] Step S104: for any two loss points, the minimum value of the two values ​​obtained by multiplying the vertical coordinates of the two loss points by the loss ratio is recorded as the loss threshold; when the difference between the vertical coordinates of the two loss points is less than the loss threshold, the horizontal coordinates of the two loss points are marked as the first-level comparison data and the second-level comparison data; when the core hidden dangers and coordinated hidden dangers of the first-level comparison data are the same as the core hidden dangers and coordinated hidden dangers of the second-level comparison data, the core hidden dangers and coordinated hidden dangers of the first-level comparison data are recorded as the financial risk combination;

[0047] Step S105, when the core risks and coordinated risks of the primary comparison data are only partially identical to the core risks and coordinated risks of the secondary comparison data, the core risks and coordinated risks of the primary comparison data and the core risks and coordinated risks of the secondary comparison data that are identical to each other are recorded as a financial risk combination;

[0048] In the specific implementation process, for example, in a data processing, the primary comparison data and the secondary comparison data are clothing export data and jewelry export data respectively, and the core hidden danger of clothing export data is market exchange rate risk, and the coordinated hidden danger of clothing export data is market interest rate risk and product pricing risk; the core hidden danger of jewelry export data is product pricing risk, and the coordinated hidden danger of jewelry export data is product pricing risk; then through analysis, it can be obtained that the core hidden danger of clothing export data is different from the core hidden danger of jewelry export data, and the coordinated hidden danger of clothing export data and the coordinated hidden danger of jewelry export data are only partially the same, then the core hidden danger and the coordinated hidden danger of clothing export data can be recorded as one financial risk combination, and the core hidden danger and the coordinated hidden danger of jewelry export data can be recorded as another financial risk combination;

[0049] Step S106, when the core hidden dangers and coordinated hidden dangers of the primary comparison data are completely different from the core hidden dangers and coordinated hidden dangers of the secondary comparison data, the core hidden dangers and coordinated hidden dangers of the primary comparison data and the core hidden dangers and coordinated hidden dangers of the secondary comparison data are recorded as two financial risk combinations respectively;

[0050] Step S107, for any risk reference data that is not recorded as primary comparison data or secondary comparison data, the core hidden danger and the coordinated hidden danger corresponding to the risk reference data are recorded as a risk financial combination;

[0051] Step S108, obtaining all risk financial combinations corresponding to the financial risk hazards; obtaining all risk financial combinations corresponding to all financial risk hazards.

[0052] Step S2, using a combined weight analysis method to analyze all financial risk combinations, and obtaining multiple risk weights for each financial risk hazard based on the analysis results; and using a weighted comprehensive algorithm based on the risk weights to obtain a comprehensive weight corresponding to each financial risk hazard;

[0053] Step S201, the combination weight analysis method includes: for any risk financial combination corresponding to all financial risk hazards, setting the risk weight of the core risk in the financial risk combination to 1, recording the number of coordinated risks in the financial risk combination as m, and setting the risk weight of any coordinated risk in the financial risk combination to 1 / m; obtaining the risk weights corresponding to the core risk and the coordinated risk in all risk financial combinations corresponding to the financial risk hazards, wherein the same financial risk hazard can obtain multiple independent risk weights in multiple financial risk combinations;

[0054] Step S202: for any financial risk hazard: the weighted comprehensive algorithm is: Among them, F is the comprehensive weight, M is the number of financial risk portfolios, and f iis the risk weight of the financial risk hidden danger in each financial risk combination, wherein, when the financial risk hidden danger does not obtain the risk weight in the financial risk combination, the risk weight of the financial risk hidden danger in the financial risk combination is set to 0; in the specific implementation process, for example, in a data processing, the financial risk hidden danger analyzed is the market exchange rate risk, and the number of financial risk combinations is 10, and the risk weights of the financial risk hidden danger in each financial risk combination are |1, 1, 0, 0, 0, 0, 0.2, 0.4, 0.1 and 1 respectively, then it can be obtained by calculation that the comprehensive weight corresponding to the market exchange rate risk is 0.185;

[0055] Step S203, obtaining the comprehensive weights corresponding to all financial risk hazards.

[0056] Step S3, obtaining a basic risk control model based on big data, and training the basic risk control model using all financial risk hazards and corresponding comprehensive weights, and obtaining a financial risk control model based on the training results;

[0057] Step S301, obtaining a basic risk control model based on big data; inputting all financial risk hazards into the basic risk control model in sequence for recording; in the basic risk control model: when the comprehensive weight of the newly recorded financial risk hazard is greater than the maximum value of the comprehensive weight of the recorded financial risk hazard, the rating of the financial risk hazard with the highest risk rating at this time is lowered, and the risk rating of the newly recorded financial risk hazard is set to the highest; when the comprehensive weight of the newly recorded financial risk hazard is less than the minimum value of the comprehensive weight of the recorded financial risk hazard, the risk rating of the financial risk hazard with the lowest risk rating at this time is increased, and the risk rating of the newly recorded financial risk hazard is set to the lowest;

[0058] Step S302: when the comprehensive weight of the newly recorded financial risk hazard is less than the maximum value of the comprehensive weights of the recorded financial risk hazard and greater than the minimum value of the comprehensive weights of the recorded financial risk hazard, the risk rating of the newly recorded financial risk hazard is set to the middle value of all risk ratings;

[0059] In the specific implementation process, by setting the rating of recorded financial risk hazards based on comprehensive weights, it can ensure that under any combination of financial risks, the losses caused by each risk hazard can still be effectively analyzed, preventing users from still having reduced returns or even losses after risk control;

[0060] Step S303: After all financial risk hazards are input into the basic risk control model, the basic risk control model at this time is recorded as the financial risk control model.

[0061] Example 2, please refer to Figure 3 As shown, Figure 3The structural schematic diagram of an electronic device is illustrated, and the electronic device may include: a processor, a communication interface, a memory and a communication bus, wherein the processor, the communication interface and the memory communicate with each other through the communication bus. The memory stores computer-readable instructions, and the processor can call the instructions in the memory. When the computer-readable instructions are executed by the processor, the steps in the risk control model construction method based on big data are executed to achieve the following functions: first, multiple financial risk hazards and risk reference data are obtained based on big data; the risk reference data is analyzed, and the financial risk combination is obtained based on the analysis results; then all financial risk combinations are analyzed using the combined weight analysis method, and multiple risk weights of each financial risk hazard are obtained based on the analysis results, and the comprehensive weight corresponding to each financial risk hazard is obtained by using a weight comprehensive algorithm based on the risk weight; finally, a basic risk control model is obtained based on big data, and the basic risk control model is trained using all financial risk hazards and the corresponding comprehensive weights, and a financial risk control model is obtained based on the training results.

[0062] In addition, the logic instructions in the above-mentioned memory can be implemented in the form of software functional units and can be stored in a computer-readable storage medium when sold or used as an independent product. Based on this understanding, the technical solution of the present application can be essentially or partly embodied in the form of a software product that contributes to the prior art, and the computer software product is stored in a storage medium, including several instructions to enable a computer device (which can be a personal computer, a server, or a network device, etc.) to perform all or part of the steps of the method described in each embodiment of the present application. The aforementioned storage medium includes: various media that can store program codes, such as a USB flash drive, a mobile hard disk, a read-only memory (ROM), a random access memory (RAM), a disk or an optical disk.

[0063] Embodiment 3. The present application also provides a computer-readable storage medium. The present application provides a storage medium on which a computer program is stored. When the computer program is executed by a processor, the steps in the risk control model construction method based on big data are executed to achieve the following functions: first, multiple financial risk hazards and risk reference data are obtained based on big data; the risk reference data is analyzed, and a financial risk combination is obtained based on the analysis results; then all financial risk combinations are analyzed using a combination weight analysis method, and multiple risk weights for each financial risk hazard are obtained based on the analysis results, and a comprehensive weight corresponding to each financial risk hazard is obtained by using a weight comprehensive algorithm based on the risk weight; finally, a basic risk control model is obtained based on big data, and the basic risk control model is trained using all financial risk hazards and the corresponding comprehensive weights, and a financial risk control model is obtained based on the training results.

[0064] Through the description of the above implementation methods, the embodiments of the present invention can be provided as methods, systems or computer program products. Based on such an understanding, the above technical solutions can be essentially or partly contributed to the prior art in the form of software products, which can be stored in a computer-readable storage medium, such as ROM / RAM, a magnetic disk, an optical disk, etc., and include several instructions for enabling a computer device (which can be a personal computer, a server, or a network device, etc.) to execute the methods described in each embodiment or some parts of the embodiments.

[0065] In the embodiments provided in the present application, it should be understood that the disclosed system or method can be implemented in other ways. The embodiments described above are merely schematic. For example, the division of modules or units is only a logical function division. There may be other division methods in actual implementation. For example, multiple modules or units can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some communication interfaces, and the indirect coupling or communication connection of systems, modules and units can be electrical, mechanical or other forms.

[0066] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present application, rather than to limit it. Although the present application has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present application.

Claims

1. A method for constructing a risk control model based on big data, characterized in that: The steps include: Obtain multiple financial risk hazards based on big data, and obtain risk reference data based on multiple financial risk hazards; analyze the risk reference data, and obtain financial risk combinations based on the analysis results; Use the combined weight analysis method to analyze all financial risk combinations and obtain multiple risk weights for each financial risk potential based on the analysis results; Based on the risk weight, a weighted comprehensive algorithm is used to obtain the comprehensive weight corresponding to each financial risk hazard; Obtain the basic risk control model based on big data, use all financial risk hazards and corresponding comprehensive weights to train the basic risk control model, and obtain the financial risk control model based on the training results.

2. The method for constructing a risk control model based on big data according to claim 1, characterized in that: Obtain multiple financial risk hazards based on big data, and obtain risk reference data based on multiple financial risk hazards, including: Based on big data, multiple hidden risks in the financial industry are obtained and recorded as financial hidden risks; for any financial hidden risk, multiple financial data containing financial hidden risks on the Internet are obtained based on big data and recorded as risk reference data; for any risk reference data, the financial hidden risks in the risk reference data are recorded as core hidden risks; the hidden risks other than financial hidden risks in the risk reference data are recorded as collaborative hidden risks; Obtain the core hidden dangers and coordinated hidden dangers of all risk reference data corresponding to the financial risk hidden dangers; obtain the loss amount in all risk reference data of the financial risk hidden dangers; establish a plane rectangular coordinate system, recorded as the hidden danger analysis coordinate system, wherein the unit of the Y-axis of the hidden danger analysis coordinate system is the amount, and the coordinate points in the X-axis from the coordinate origin to the right are filled in with all the risk reference data corresponding to the financial risk hidden dangers in sequence.

3. The method for constructing a risk control model based on big data according to claim 2, characterized in that: Analyze the risk reference data and obtain the financial risk portfolio based on the analysis results, including: Based on the loss amount of each risk reference data, mark the points in the hidden danger analysis coordinate system and record them as loss points; record the number of risk reference data as n, and record (1 / n)×100% as the loss ratio; For any two loss points, the minimum value of the two values ​​obtained by multiplying the vertical coordinates of the two loss points by the loss ratio is recorded as the loss threshold; when the difference between the vertical coordinates of the two loss points is less than the loss threshold, the horizontal coordinates of the two loss points are marked as primary comparison data and secondary comparison data; when the core hidden dangers and coordinated hidden dangers of the primary comparison data are the same as the core hidden dangers and coordinated hidden dangers of the secondary comparison data, the core hidden dangers and coordinated hidden dangers of the primary comparison data are recorded as financial risk combinations.

4. The method for constructing a risk control model based on big data according to claim 3 is characterized in that: Analyzing the risk reference data and obtaining the financial risk portfolio based on the analysis results also includes: When the core hidden dangers and coordinated hidden dangers of the primary comparison data are only partially identical with the core hidden dangers and coordinated hidden dangers of the secondary comparison data, the core hidden dangers and coordinated hidden dangers of the primary comparison data and the core hidden dangers and coordinated hidden dangers of the secondary comparison data that are identical to each other shall be recorded as a financial risk combination; When the core hidden dangers and coordinated hidden dangers of the primary comparison data are completely different from the core hidden dangers and coordinated hidden dangers of the secondary comparison data, the core hidden dangers and coordinated hidden dangers of the primary comparison data and the core hidden dangers and coordinated hidden dangers of the secondary comparison data shall be recorded as two financial risk combinations respectively.

5. The method for constructing a risk control model based on big data according to claim 4, characterized in that: Analyzing the risk reference data and obtaining the financial risk portfolio based on the analysis results also includes: For any risk reference data that is not recorded as primary comparison data or secondary comparison data, the core hidden dangers and coordinated hidden dangers corresponding to the risk reference data are recorded as risk financial combinations; Obtain all risky financial combinations corresponding to financial risk hazards; obtain all risky financial combinations corresponding to all financial risk hazards.

6. The method for constructing a risk control model based on big data according to claim 5, characterized in that: The combined weight analysis method includes: For any risk financial combination corresponding to all financial risk hazards, the risk weight of the core hazard in the financial risk combination is set to 1, the number of collaborative hazards in the financial risk combination is recorded as m, and the risk weight of any collaborative hazard in the financial risk combination is set to 1 / m; obtain the risk weights corresponding to the core hazards and collaborative hazards in all risk financial combinations corresponding to the financial risk hazards, where the same financial risk hazard can obtain multiple independent risk weights in multiple financial risk combinations.

7. The method for constructing a risk control model based on big data according to claim 6, characterized in that: Based on the risk weight, the weighted comprehensive algorithm is used to obtain the comprehensive weight corresponding to each financial risk hazard, including: For any financial risk: the weighted comprehensive algorithm is: Among them, F is the comprehensive weight, M is the number of financial risk portfolios, and f i is the risk weight of the financial risk hazard in each financial risk portfolio, wherein, when the financial risk hazard does not obtain a risk weight in the financial risk portfolio, the risk weight of the financial risk hazard in the financial risk portfolio is set to 0; Obtain the comprehensive weights corresponding to all financial risk hazards.

8. The method for constructing a risk control model based on big data according to claim 7, characterized in that: Based on big data, the basic risk control model is obtained, and all financial risk hazards and corresponding comprehensive weights are used to train the basic risk control model. The financial risk control model obtained based on the training results includes: A basic risk control model is obtained based on big data; all financial risk hazards are input into the basic risk control model in sequence for recording; in the basic risk control model: when the comprehensive weight of a newly recorded financial risk hazard is greater than the maximum value of the comprehensive weights of the recorded financial risk hazards, the rating of the financial risk hazard with the highest risk rating at this time is lowered, and the risk rating of the newly recorded financial risk hazard is set to the highest; when the comprehensive weight of a newly recorded financial risk hazard is less than the minimum value of the comprehensive weights of the recorded financial risk hazards, the risk rating of the financial risk hazard with the lowest risk rating at this time is increased, and the risk rating of the newly recorded financial risk hazard is set to the lowest.

9. The method for constructing a risk control model based on big data according to claim 8, characterized in that: Obtaining a basic risk control model based on big data, and using all financial risk hazards and corresponding comprehensive weights to train the basic risk control model. Obtaining a financial risk control model based on the training results also includes: When the comprehensive weight of a newly recorded financial risk hazard is less than the maximum value of the comprehensive weights of the recorded financial risk hazards and greater than the minimum value of the comprehensive weights of the recorded financial risk hazards, the risk rating of the newly recorded financial risk hazard is set to the middle value of all risk ratings.

10. The method for building a risk control model based on big data according to claim 9, characterized in that: Obtaining a basic risk control model based on big data, and using all financial risk hazards and corresponding comprehensive weights to train the basic risk control model. Obtaining a financial risk control model based on the training results also includes: When all financial risk hazards are input into the basic risk control model, the basic risk control model at this time will be recorded as the financial risk control model.

Images