A display system for underwriting result data

Abstract

The embodiment of the present invention provides a display system for underwriting result data, including a data collection module, an underwriting result calculation module, and a display module. The underwriting result calculation module includes an import submodule for importing underwriting feature data into multiple preset trainers; an initial underwriting result data calculation submodule for using the initial weights and underwriting feature data of the trainer to calculate the initial underwriting result data of the trainer; a residual calculation submodule for using the initial underwriting result data to calculate the residual of the trainer; a target weight calculation submodule for using the residual to calculate the target weight of the trainer; an intermediate underwriting result data calculation submodule for using the target weight and underwriting feature data to calculate the intermediate underwriting result data of the trainer; and a target underwriting result data calculation submodule for using the intermediate underwriting result data to calculate the target underwriting result data. Through the embodiment of the present invention, the accuracy of the underwriting result data is improved.

Description

Technical Field

[0001] The present invention relates to the technical field of data display, and in particular to a display system for underwriting result data. Background Art

[0002] The traditional underwriting process mainly involves professional underwriters reading the insurance applicant's physical examination report and some auxiliary materials to determine whether the insurance company should sell the insurance to the applicant.

[0003] In recent years, intelligent underwriting or automated underwriting has gradually emerged in the field of insurance underwriting. There are two main difficulties in realizing intelligent underwriting: one is the recognition and structuring of physical examination reports, and the other is the generation and display of underwriting result data.

[0004] The traditional approach of intelligent underwriting generally uses XGBoost to generate underwriting result data. However, in actual underwriting scenarios, the accuracy of underwriting result data predicted by XGBoost is not satisfactory. In the process of generating underwriting result data, it may not capture some data features, such as the abstract relationship and interaction of some variables. Summary of the invention

[0005] In view of the above problems, an embodiment of the present invention is proposed to provide a system for displaying underwriting result data that overcomes the above problems or at least partially solves the above problems.

[0006] In order to solve the above problems, an embodiment of the present invention discloses a system for displaying underwriting result data, including:

[0007] Data collection module, underwriting result calculation module, and display module;

[0008] The data collection module is used to collect underwriting feature data;

[0009] The underwriting result calculation module includes:

[0010] An import submodule, used to import the underwriting feature data into a plurality of preset trainers; each trainer has a corresponding initial weight;

[0011] An initial underwriting result data calculation submodule, used to calculate the initial underwriting result data of the trainer using the initial weight of the trainer and the underwriting feature data;

[0012] A residual calculation submodule, used to calculate the residual of the trainer using the initial underwriting result data;

[0013] A target weight calculation submodule, used to calculate the target weight of the trainer using the residual;

[0014] An intermediate underwriting result data calculation submodule, used to calculate the intermediate underwriting result data of the trainer using the target weight and the underwriting feature data;

[0015] A target underwriting result data calculation submodule, used to calculate the target underwriting result data using the intermediate underwriting result data;

[0016] The display module is used to display the target underwriting result data.

[0017] Optionally, the trainers have corresponding numbers, and the residual includes the residual of the first trainer, wherein the first trainer is the trainer with the smallest number; the residual calculation submodule includes:

[0018] A first underwriting result data extraction unit, configured to extract first initial underwriting result data of the first trainer;

[0019] The first trainer residual calculation unit is used to calculate the residual between the first initial underwriting result data and the preset standard underwriting result data as the residual of the first trainer.

[0020] Optionally, the residual also includes a residual of a second trainer, wherein the second trainer is a trainer other than the trainer with the smallest number; and the residual calculation submodule further includes:

[0021] A traversal unit, used for traversing the second trainer according to the number;

[0022] a second initial underwriting result data extraction unit, configured to determine a current trainer in the second trainer and extract second initial underwriting result data of the current trainer;

[0023] The residual extraction unit is used to extract the residual of the previous trainer;

[0024] a current trainer residual calculation unit, configured to calculate a residual of the previous trainer and a residual of the second initial underwriting result data as a residual of the current trainer;

[0025] A residual writing unit, used for writing the residual of the current trainer into the residual of the second trainer;

[0026] A traversal judgment unit is used to judge whether all numbers have been traversed;

[0027] A residual saving unit, used for saving the residual of the second trainer if yes;

[0028] A returning unit is used for returning to the step of determining the current trainer in the second trainer if no.

[0029] Optionally, the target weight includes a first target weight; and the steps of the target weight calculation submodule include:

[0030] A first loss function generating unit, used to generate a first loss function of the trainer, wherein the first loss function is a square value of a residual of the trainer;

[0031] A first penalty variable function acquisition unit, used to acquire a first penalty variable function of the trainer;

[0032] The first target weight determination unit is used to use the quotient of the first penalty variable function and a preset first hyperparameter as the first target weight of the trainer when the sum of the first loss function and the first penalty variable function is minimized.

[0033] Optionally, the target underwriting result data includes first target underwriting result data; and the target underwriting result data calculation submodule includes:

[0034] An intermediate underwriting result data acquisition unit, used to acquire the intermediate underwriting result data within a preset threshold range;

[0035] The first target underwriting result data determining unit is used to sum the intermediate underwriting result data within a preset threshold range as the first target underwriting result data.

[0036] Optionally, the target weight includes a second target weight; and the target weight calculation submodule includes:

[0037] A second loss function generating unit, used to generate a second loss function of the trainer, where the second loss function is the sum of squares of residuals of all the trainers;

[0038] A second penalty variable function acquisition unit, used to acquire a second penalty variable function of the trainer;

[0039] The second target weight determination unit is used to use the quotient of the second penalty variable function and a preset second hyperparameter as the second target weight of the trainer when the sum of the second loss function and the second penalty variable function is minimized.

[0040] Optionally, the target underwriting result data includes second target underwriting result data; and the target underwriting result data calculation submodule includes:

[0041] The second target underwriting result data determining unit is configured to sum up all the intermediate underwriting result data as the second target underwriting result data. Optionally, the first target weight determining unit includes:

[0042] The first target weight determination subunit is used to use the quotient of the first penalty variable function and a preset first hyperparameter as the first target weight of the trainer.

[0043] Optionally, the second target weight determination unit includes:

[0044] The second target weight determination subunit is used to use the quotient of the second penalty variable function and a preset second hyperparameter as the second target weight of the trainer.

[0045] The embodiment of the present invention further discloses an electronic device, including:

[0046] A system for displaying underwriting result data as described in any of the above items.

[0047] The embodiment of the present invention includes the following advantages: The embodiment of the present invention proposes a display system for underwriting result data. By acquiring underwriting feature data; importing the underwriting feature data into multiple preset trainers; using the initial weights and underwriting feature data of the trainer to calculate the initial underwriting result data of the trainer; using the initial underwriting result data to calculate the residual of the trainer; using the residual to calculate the target weight of the trainer; using the target weight and underwriting feature data to calculate the intermediate underwriting result data of the trainer; using the intermediate underwriting result data to calculate the target underwriting result data and display the target underwriting data. The underwriting result data output by the trainer is close to the true value, which improves the accuracy of the underwriting result data. BRIEF DESCRIPTION OF THE DRAWINGS

[0048] Figure 1 It is a structural block diagram of a display system for underwriting result data of the present invention;

[0049] Figure 2 It is a flowchart of the steps of a method for a trainer to output underwriting result data in an embodiment of the present invention. DETAILED DESCRIPTION

[0050] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and easy to understand, the present invention is further described in detail below with reference to the accompanying drawings and specific embodiments.

[0051] One of the core concepts of an embodiment of the present invention is that, based on the residual cascade method, the underwriting result data output by the current trainer is continuously used to fit the residuals of the previous trainer, and the weights of the trainer are optimized based on the residuals, so that the underwriting result data output by the trainer is closer to the true value, thereby improving the accuracy of the underwriting result data finally output.

[0052] Reference Figure 1 , shows a structural block diagram of a display system for underwriting result data of the present invention, which may specifically include the following modules:

[0053] Data collection module 101, used to collect underwriting feature data;

[0054] In an embodiment of the present invention, the underwriting characteristic data is data related to the generation of underwriting result data extracted based on the applicant's physical examination report and some auxiliary materials, such as the physical examination items in the physical examination report and the corresponding physical examination results (such as "height: 170cm").

[0055] The data collection module collects underwriting feature data and analyzes the underwriting feature data to determine whether to underwrite the applicant.

[0056] The underwriting result calculation module 102 may include the following submodules:

[0057] An import submodule 11 is used to import the underwriting feature data into a plurality of preset trainers; each trainer has a corresponding initial weight;

[0058] A trainer is used to generate underwriting result data using input underwriting feature data.

[0059] Each trainer has a corresponding initial weight; the initial weight is a weight pre-set for each trainer, and the initial weight includes the weight of each underwriting feature data.

[0060] In the embodiment of the present invention, the weight of each underwriting feature data is different. In actual application, a higher weight can be set for the underwriting feature data that has a greater correlation with the generation of underwriting result data.

[0061] In an embodiment of the present invention, the import submodule can import the underwriting feature data into multiple preset trainers and output different results according to different initial weights.

[0062] The initial underwriting result data calculation submodule 12 is used to calculate the initial underwriting result data of the trainer using the initial weight of the trainer and the underwriting feature data.

[0063] In an embodiment of the present invention, the initial underwriting result data calculation submodule may use the initial weights of the trainer and the input underwriting feature data to calculate the initial underwriting result data of the trainer.

[0064] The residual calculation submodule 13 is used to calculate the residual of the trainer using the initial underwriting result data.

[0065] In this embodiment of the present invention, the residual calculation submodule 13 may include the following units:

[0066] A first underwriting result data extraction unit 1301, configured to extract first initial underwriting result data of the first trainer;

[0067] Since the initial underwriting result data of the trainer is calculated based on the initial weight, and the initial weight is a pre-set value, it may deviate from the actual value, resulting in inaccurate output results. Therefore, the embodiment of the present invention can improve the accuracy of the output results by optimizing the initial weight.

[0068] In the embodiment of the present invention, the trainers have corresponding numbers. Therefore, first, the first initial underwriting result data extraction unit can be used to extract the first initial underwriting result data of the first trainer. The first trainer is the trainer with the smallest number.

[0069] A first trainer residual calculation unit 1302, configured to calculate a residual between the first initial underwriting result data and preset standard underwriting result data as a residual of the first trainer;

[0070] After obtaining the first initial underwriting result data, the first trainer residual calculation unit can calculate the residual between the first initial underwriting result data and the preset standard underwriting result data as the residual of the first trainer. Specifically, the residual e1 of the first trainer can be expressed as:

[0071]

[0072] Wherein, y is the preset standard underwriting result data, which represents the actual underwriting result data;

[0073] This is the first initial underwriting result data.

[0074] A traversal unit 1303, configured to traverse the second trainer according to the number;

[0075] Since the residuals of trainers other than the first trainer are obtained in the same way, the trainers other than the first trainer can be classified as second trainers, and a traversal unit is used to traverse each trainer in the second trainer according to the number corresponding to each trainer.

[0076] A second initial underwriting result data extraction unit 1304, configured to determine a current trainer in the second trainer and extract second initial underwriting result data of the current trainer;

[0077] In the embodiment of the present invention, the second initial underwriting result data extraction unit can determine the current trainer in the second trainer and extract the second initial underwriting result data of the current trainer.

[0078] A residual extraction unit 1305, used to extract the residual of the previous trainer;

[0079] In the embodiment of the present invention, the residual of the previous trainer can be extracted by the residual extraction unit and used as the fitting value of the current trainer. For example, if the trainer numbered 2 is determined to be the current trainer, the residual of the trainer numbered 1 is extracted. And so on.

[0080] A current trainer residual calculation unit 1306 is used to calculate the residual of the previous trainer and the residual of the second initial underwriting result data as the residual of the current trainer;

[0081] In the embodiment of the present invention, the second initial underwriting result data output by the current trainer is used to fit the residual of the previous trainer. Therefore, the residual of the current trainer is the residual between the second initial underwriting result data of the current trainer and the residual of the previous trainer.

[0082] For example, when the current trainer is trainer number 2, its output is is to fit the residual of the first trainer. Therefore, the residual e2 of the current trainer can be expressed as:

[0083]

[0084] Similarly, the output of the i-th trainer is to fit the residual e i-1 , then the residual e of the trainer numbered i i It can be expressed as:

[0085]

[0086] A residual writing unit 1307, used to write the residual of the current trainer into the residual of the second trainer;

[0087] The residual writing unit may write the residual of the current trainer into the residual of the second trainer after acquiring the residual of the current trainer.

[0088] A traversal determination unit 1308 is used to determine whether all numbers have been traversed;

[0089] The traversal judgment unit can judge whether all numbers have been traversed, that is, whether all trainers have obtained the corresponding residuals.

[0090] A residual saving unit 1309 is used to save the residual of the second trainer if yes;

[0091] After all the second trainers have obtained the corresponding residuals, the obtained residuals can be saved by the residual saving unit.

[0092] The returning unit 1310 is configured to return to the step of determining the current trainer in the second trainer if no;

[0093] When all the numbers have not been traversed, the returning unit may be used to return to the process of determining the current trainer in the second trainer, until the residuals of all the second trainers are obtained.

[0094] A target weight calculation submodule 14 is used to calculate the target weight of the trainer using the residual;

[0095] In the embodiment of the present invention, after the residual of each trainer is obtained, the target weight calculation submodule can use the residual to calculate the target weight of each trainer.

[0096] In one example, a cyclic training mechanism may be used to train each trainer, thereby optimizing the weight of each trainer to obtain a target weight, and then optimizing the output result of each trainer through the target weight.

[0097] To this end, according to the number of trainers, it is necessary to first design a corresponding number of objective functions (assuming N), so as to determine the target weights based on the objective functions.

[0098] In the embodiment of the present invention, the target weight calculation submodule 14 may specifically include the following units:

[0099] A first loss function generating unit 1411 is used to generate a first loss function of the trainer, where the first loss function is a square value of a residual of the trainer;

[0100] A loss function is a function that measures the extent of a loss associated with an "erroneous" estimate, such as loss of money or equipment.

[0101] The first loss function is a component of the objective function. In the embodiment of the present invention, the first loss function generating unit can use the square value of the residual of each trainer as the first loss function. Specifically, it can be expressed as:

[0102]

[0103] A first penalty variable function acquisition unit 1412, used to acquire a first penalty variable function of the trainer;

[0104] The first penalty variable function is the penalty term of the objective function, which is used to constrain the optimization space.

[0105] In the embodiment of the present invention, the first penalty variable function acquisition unit may acquire the first penalty variable function of the trainer.

[0106] In the embodiment of the present invention, the first penalty variable function can be expressed as:

[0107] λ||ωi ||2

[0108] Among them, λ is the hyperparameter of the first penalty variable function.

[0109] ||ω i ||2 is the target weight of the i-th trainer.

[0110] After obtaining the first penalty variable function and the first loss function, the objective function can be obtained:

[0111]

[0112] Since the N trainers are trained in a loop, the optimization of the N objective functions is performed in sequence. The objective function for the next network module unit is a constant. Therefore, in the objective function, the variable is and ||ω i ||2, where Yes||ω i ||2 dependent variable. By adjusting the variable to and ||ω i ||2, the value of the objective function can be adjusted.

[0113] The first target weight determining unit 1413 is configured to generate a first target weight by using the first penalty variable function when the sum of the first loss function and the first penalty variable function is minimized.

[0114] In an embodiment of the present invention, the first target weight determination unit may include a first target weight determination subunit, configured to use the quotient of the first penalty variable function and a preset first hyperparameter as the first target weight of the trainer.

[0115] In the embodiment of the present invention, by adjusting ||ω i ||2 makes the value of the objective function minimum, that is, the sum of the first loss function and the first penalty variable function is minimized to determine the first objective weight. When the value of the objective function is minimum, ||ω i ||2 is the first target weight of the i-th trainer.

[0116] In one embodiment of the present invention, a simultaneous training mechanism can also be used to train all trainers. The simultaneous training mechanism only needs to define a total objective function, and only the total objective function needs to be optimized during training; through the total objective function, the target weight of each trainer can be obtained.

[0117] Under the simultaneous training mechanism, the target weight calculation submodule 14 may include the following units:

[0118] A second loss function generating unit 1421 is used to generate a second loss function of the trainer, where the second loss function is the sum of squares of the residuals of all the trainers;

[0119] In an embodiment of the present invention, the second loss function generating unit may use the sum of squares of the residuals of all trainers as the loss function of the objective function.

[0120] A second penalty variable function acquisition unit 1422, used to acquire a second penalty variable function of the trainer;

[0121] In the embodiment of the present invention, the second penalty variable function of the trainer may be acquired by the second penalty variable function acquisition unit; the second penalty variable function is the sum of the penalty variable functions of each trainer.

[0122] The second target weight determining unit 1423 is configured to generate a second target weight by using the second penalty variable function when the sum of the second loss function and the second penalty variable function is minimized.

[0123] In an embodiment of the present invention, the second target weight determination unit may include a second target weight determination subunit, and the second target weight determination subunit may use the quotient of the second penalty variable function and the preset second hyperparameter as the second target weight of the trainer when the value of the second target function is minimum, that is, the sum of the second loss function and the second penalty variable function is minimum. The second target weight is a weight set, which includes the second target weight of each trainer. According to different numbers, the corresponding second target weight can be obtained.

[0124] An intermediate underwriting result data calculation submodule 15, configured to calculate the intermediate underwriting result data of the trainer using the target weight and the underwriting feature data;

[0125] In an embodiment of the present invention, the intermediate underwriting result data calculation submodule can calculate the intermediate underwriting result data of the trainer according to the target weight and the underwriting feature data after obtaining the target weight.

[0126] The target underwriting result data calculation submodule 16 is used to calculate the target underwriting result data using the intermediate underwriting result data.

[0127] In an embodiment of the present invention, after obtaining the intermediate underwriting result data output by the trainer, the target underwriting result data calculation submodule can use the intermediate underwriting result data to calculate the target underwriting result data.

[0128] In one example, the target underwriting result data calculation submodule 16 may include the following units:

[0129] The intermediate underwriting result data acquisition unit 1611 is used to acquire the intermediate underwriting result data within a preset threshold range;

[0130] The first target underwriting result data determining unit 1612 is configured to sum the intermediate underwriting result data within a preset threshold range as the first target underwriting result data.

[0131] When a cyclic training mechanism is selected to optimize the trainer, in actual training, after a certain number of trainers are cycled (assuming T, T≤N), the output of the subsequent trainer will become very small, far less than the output value of the previous trainer, and is completely insufficient to affect the output of the final target underwriting result data. Therefore, in an embodiment of the present application, the trainer can be pruned to retain the trainer that has a greater impact on the target underwriting result data. The sum of the intermediate underwriting result data of the retained trainers is used as the target underwriting result data. This shortens the time for calculating the target underwriting result data.

[0132] In one example, the target underwriting result data calculation submodule 16 may include the following units:

[0133] The second target underwriting result data determining unit 1621 is used to sum up all the intermediate underwriting result data as the second target underwriting result data. Specifically, it can be expressed as the following formula:

[0134]

[0135] Reference Figure 2 , showing a flow chart of the steps of a trainer outputting underwriting result data according to an embodiment of the present invention.

[0136] like Figure 2 As shown, 55 underwriting feature data (each underwriting feature data is 1-dimensional) are input to the trainer. After receiving the 55 underwriting feature data, the trainer will convert each underwriting feature data into 7-dimensional feature data in the form of entity embedding (201); then all the underwriting feature data are expanded (202) to obtain 385-dimensional feature data; the 385-dimensional feature data is batch normalized and randomly inactivated by the ReLU activation function, so that part of the feature data can be randomly removed; then based on full connection (203), according to the weight of each feature data and the weight of the neurons in the previous layer, the remaining feature data can be converted into the 8 neurons in the upper layer; then based on full connection (204), according to the weight of the data in the 8 neurons and the weight of the neurons in the next layer, the data in the 8 neurons can be converted into the values ​​in the four neurons in the next layer. The value group composed of the values ​​of these four neurons is the underwriting result data output by the trainer.

[0137] The display module 103 is used to display the target underwriting result data.

[0138] In one embodiment of the present invention, the display module can be used to display target underwriting result data, and the displayed data is derived from the underwriting result calculation module, and what is displayed is the target underwriting result data.

[0139] In an embodiment of the present invention, the underwriting result data display system can obtain the residual of the trainer and use the residual to optimize the weight of the trainer, so that the underwriting result data output by the trainer is close to the true value, thereby improving the accuracy of the underwriting result data.

[0140] It should be noted that, for the sake of simplicity, the system embodiments are described as a series of action combinations, but those skilled in the art should be aware that the embodiments of the present invention are not limited by the described action sequence, because according to the embodiments of the present invention, certain steps can be performed in other sequences or simultaneously. Secondly, those skilled in the art should also be aware that the embodiments described in the specification are all preferred embodiments, and the actions involved are not necessarily required by the embodiments of the present invention.

[0141] The embodiment of the present invention further discloses an electronic device, including:

[0142] A system for displaying underwriting result data as described in the above embodiments of the present application.

[0143] The various embodiments in this specification are described in a progressive manner, and each embodiment focuses on the differences from other embodiments. The same or similar parts between the various embodiments can be referenced to each other.

[0144] It will be appreciated by those skilled in the art that the embodiments of the present invention may be provided as methods, devices, or computer program products. Therefore, the embodiments of the present invention may take the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware. Moreover, the embodiments of the present invention may take 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.) containing computer-usable program codes.

[0145] The embodiments of the present invention are described with reference to the flowcharts and / or block diagrams of the methods, terminal devices (systems), and computer program products according to the embodiments of the present invention. It should be understood that each process and / or block in the flowchart and / or block diagram, as well as the combination of the processes and / or blocks 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 terminal device to generate a machine, so that the instructions executed by the processor of the computer or other programmable data processing terminal 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.

[0146] These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing terminal device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce a manufactured product including 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.

[0147] These computer program instructions can also be loaded onto a computer or other programmable data processing terminal device so that a series of operating steps are executed on the computer or other programmable terminal device to produce a computer-implemented process, thereby providing instructions for executing on the computer or other programmable terminal device to implement 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.

[0148] Although the preferred embodiments of the present invention have been described, those skilled in the art may make additional changes and modifications to these embodiments once they have learned the basic creative concept. Therefore, the appended claims are intended to be interpreted as including the preferred embodiments and all changes and modifications that fall within the scope of the embodiments of the present invention.

[0149] Finally, it should be noted that, in this article, relational terms such as first and second, etc. are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "include", "comprise" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or terminal 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 terminal device. In the absence of further restrictions, the elements defined by the sentence "comprise a ..." do not exclude the existence of other identical elements in the process, method, article or terminal device including the elements.

[0150] The above is a detailed introduction to a display system for underwriting result data provided by the present invention. Specific examples are used herein to illustrate the principles and implementation methods of the present invention. The description of the above embodiments is only used to help understand the system of the present invention and its core ideas. At the same time, for those skilled in the art, according to the ideas of the present invention, there will be changes in the specific implementation methods and application scopes. In summary, the content of this specification should not be understood as a limitation on the present invention.

Claims

1. A display system for underwriting result data, characterized in that: include: Data collection module, underwriting result calculation module, and display module; The data collection module is used to collect underwriting feature data; The underwriting result calculation module includes: An import submodule, used for importing the underwriting feature data into a plurality of preset trainers; each trainer has a corresponding initial weight, and the trainer comprises neurons with a hierarchical structure; An initial underwriting result data calculation submodule, used to calculate the initial underwriting result data of the trainer using the initial weight of the trainer and the underwriting feature data; A residual calculation submodule, used to calculate the residual of the trainer using the initial underwriting result data; A target weight calculation submodule, used to calculate the loss function and penalty variable function of the trainer using the residual, and calculate the target weight of the trainer based on the loss function and penalty variable function of the trainer; An intermediate underwriting result data calculation submodule, used to calculate the intermediate underwriting result data of the trainer using the target weight and the underwriting feature data; A target underwriting result data calculation submodule, used to calculate the target underwriting result data using the intermediate underwriting result data; The display module is used to display the target underwriting result data.

2. The system according to claim 1, characterized in that The trainers have corresponding numbers, the residual includes the residual of the first trainer, wherein the first trainer is the trainer with the smallest number; the residual calculation submodule includes: A first underwriting result data extraction unit, configured to extract first initial underwriting result data of the first trainer; The first trainer residual calculation unit is used to calculate the residual between the first initial underwriting result data and the preset standard underwriting result data as the residual of the first trainer.

3. The system according to claim 2, characterized in that The residual also includes the residual of the second trainer, wherein the second trainer is a trainer other than the trainer with the smallest number; the residual calculation submodule also includes: A traversal unit, used for traversing the second trainer according to the number; a second initial underwriting result data extraction unit, configured to determine a current trainer in the second trainer and extract second initial underwriting result data of the current trainer; The residual extraction unit is used to extract the residual of the previous trainer; a current trainer residual calculation unit, configured to calculate a residual of the previous trainer and a residual of the second initial underwriting result data as a residual of the current trainer; A residual writing unit, used for writing the residual of the current trainer into the residual of the second trainer; A traversal judgment unit is used to judge whether all numbers have been traversed; A residual saving unit, used for saving the residual of the second trainer if yes; A returning unit is used for returning to the step of determining the current trainer in the second trainer if no.

4. The system according to claim 1, 2 or 3, characterized in that: The target weights include a first target weight; The target weight calculation submodule includes: A first loss function generating unit, used to generate a first loss function of the trainer, wherein the first loss function is a square value of a residual of the trainer; A first penalty variable function acquisition unit, used to acquire a first penalty variable function of the trainer; The first target weight determination unit is configured to generate a first target weight by using the first penalty variable function when the sum of the first loss function and the first penalty variable function is minimized.

5. The system according to claim 4, characterized in that The target underwriting result data includes first target underwriting result data; The target underwriting result data calculation submodule includes: An intermediate underwriting result data acquisition unit, used to acquire the intermediate underwriting result data within a preset threshold range; The first target underwriting result data determining unit is used to sum the intermediate underwriting result data within a preset threshold range as the first target underwriting result data.

6. The system according to claim 1, 2 or 3, characterized in that: The target weights include a second target weight; The target weight calculation submodule includes: A second loss function generating unit, used to generate a second loss function of the trainer, where the second loss function is the sum of squares of residuals of all the trainers; A second penalty variable function acquisition unit, used to acquire a second penalty variable function of the trainer; The second target weight determination unit is configured to generate a second target weight by using the second penalty variable function when the sum of the second loss function and the second penalty variable function is minimized.

7. The system according to claim 6, characterized in that The target underwriting result data includes second target underwriting result data; The target underwriting result data calculation submodule includes: The second target underwriting result data determining unit is used to add up all the intermediate underwriting result data as the second target underwriting result data.

8. The system according to claim 4, characterized in that The first target weight determination unit includes: The first target weight determination subunit is used to use the quotient of the first penalty variable function and a preset first hyperparameter as the first target weight of the trainer.

9. The system according to claim 6, characterized in that The second target weight determination unit includes: The second target weight determination subunit is used to use the quotient of the second penalty variable function and a preset second hyperparameter as the second target weight of the trainer.

10. An electronic device, characterized in that: include: A system for displaying underwriting result data as described in any one of claims 1 to 9.

Images