Insurance product pricing method and device based on artificial intelligence, equipment and medium

Abstract

The invention provides an insurance product pricing method and device based on artificial intelligence, equipment and a medium, and belongs to the field of artificial intelligence, and the method comprises the steps: obtaining health data and a health risk assessment model of an insured person; performing health risk level assessment on the health data of the insured person through a health risk assessment model to obtain a health risk level of the insured person; obtaining a basic price of the insurance product, and determining an insurance premium floating price corresponding to the insured according to the health risk level; and according to a basic price and the insurance premium floating price, determining a target price of the insurance product purchased by the insured person. According to the method, the accuracy of pricing the insurance product is greatly improved, and the satisfaction degree and recognition degree of the insured to the insurance product can be effectively improved. The method can be applied to the technical field of finance, and the accuracy of financial insurance product pricing is improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of artificial intelligence, and in particular to an insurance product pricing method and device based on artificial intelligence, equipment and medium. BACKGROUND

[0002] With the rapid development of economy, the continuous improvement of national income level and education level, the health insurance of the financial industry has gained more and more attention, including life insurance, medical insurance, disease insurance and accident insurance, etc. At present, the insurance product pricing of health insurance can only be based on age stage to carry out simple policy pricing, and cannot accurately carry out personalized product pricing for each insured person, resulting in the decrease of user's recognition of insurance products, and thus the high complaint rate and low time competitiveness of insurance products.

[0003] Therefore, how to accurately price the personalized insurance products for the insured person is a problem to be solved at present. SUMMARY

[0004] The main purpose of the present application is to provide an insurance product pricing method, device, equipment and medium based on artificial intelligence, which aims to improve the accuracy of insurance product pricing.

[0005] In a first aspect, the present application provides an insurance product pricing method based on artificial intelligence, which comprises the following steps:

[0006] Obtain the health data of the insured person and the health risk assessment model, wherein the health risk assessment model is obtained by pre-training a neural network model based on a plurality of sample data, and the sample data includes sample health data and labeled health risk level;

[0007] Evaluate the health risk level of the health data of the insured person through the health risk assessment model to obtain the health risk level of the insured person;

[0008] Obtain the basic pricing of the insurance product, and determine the premium floating price corresponding to the insured person according to the health risk level;

[0009] Determine the target pricing of the insurance product purchased by the insured person according to the basic pricing and the premium floating price.

[0010] In a second aspect, the present application further provides an insurance product pricing device, which comprises an obtaining module, a generating module and a determining module, wherein:

[0011] The acquisition module is configured to acquire health data of the insured person and a health risk assessment model, the health risk assessment model being obtained by training a neural network model based on a plurality of sample data, the sample data including sample health data and labeled health risk levels;

[0012] The generation module is configured to perform health risk level assessment on the health data of the insured person by using the health risk assessment model to obtain a health risk level of the insured person.

[0013] The acquisition module is further configured to acquire a basic pricing of an insurance product.

[0014] The determination module is configured to determine a premium floating price corresponding to the insured person according to the health risk level.

[0015] The determination module is further configured to determine a target pricing of the insurance product purchased by the insured person according to the basic pricing and the premium floating price.

[0016] In a third aspect, the present application further provides a computer device, which comprises a processor, a memory, and a computer program stored in the memory and executable by the processor, wherein the computer program, when executed by the processor, implements the steps of the artificial intelligence-based insurance product pricing method as described above.

[0017] In a fourth aspect, the present application further provides a computer readable storage medium, which stores a computer program, wherein the computer program, when executed by a processor, implements the steps of the artificial intelligence-based insurance product pricing method as described above.

[0018] The application provides an insurance product pricing method and device based on artificial intelligence, and a medium. The application obtains health data of an insured person and a health risk assessment model. The health risk assessment model is obtained by training a neural network model based on a plurality of sample data. The sample data includes sample health data and labeled health risk levels. The health risk assessment model is used to evaluate the health risk level of the health data of the insured person, and the health risk level of the insured person is obtained. The base pricing of an insurance product is obtained, and the premium floating price corresponding to the insured person is determined according to the health risk level. The target pricing of the insurance product purchased by the insured person is determined according to the base pricing and the premium floating price. In the application, the health risk level of the health data of the insured person is evaluated by the health risk assessment model, the health risk level of the insured person can be accurately obtained, the premium floating price matched by the insured person can be accurately determined based on the health risk level, and the target pricing of the insurance product purchased by the insured person can be accurately determined based on the base pricing and the premium floating price of the insurance product, thereby greatly improving the accuracy of the insurance product pricing and effectively improving the satisfaction of the insured person with the insurance product. BRIEF DESCRIPTION OF DRAWINGS

[0019] In order to more clearly illustrate the technical solutions of the embodiments of the application, the drawings needed in the embodiment description will be briefly introduced. Obviously, the drawings in the following description are some embodiments of the application, and other drawings can be obtained by those skilled in the art without creative labor.

[0020] Figure 1 A flowchart of an insurance product pricing method based on artificial intelligence provided by an embodiment of the application is shown.

[0021] Figure 2 A flowchart of another insurance product pricing method based on artificial intelligence provided by an embodiment of the application is shown.

[0022] Figure 3 A structure diagram of a health risk assessment model provided by an embodiment of the application is shown.

[0023] Figure 4 A sub-step flowchart of the insurance product pricing method based on artificial intelligence in the embodiment of the application is shown. Figure 1

[0024] A schematic block diagram of an insurance product pricing device provided by an embodiment of the application is shown. Figure 5

[0025] A schematic block diagram of a sub-module of the insurance product pricing device in the embodiment of the application is shown. Figure 6 Figure 5 ​​

[0026] Figure 7 A structural schematic block diagram of a computer device provided by an embodiment of the present application is shown.

[0027] The implementation, functional features and advantages of the present application will be further described with reference to the embodiments and the accompanying drawings. DETAILED DESCRIPTION

[0028] The technical solutions in the embodiments of the present application will be clearly and completely described with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are only some of the embodiments of the present application, but not all the embodiments of the present application. Based on the embodiments in the present application, all other embodiments obtained by those of ordinary skill in the art without creative work fall within the protection scope of the present application.

[0029] The flowchart shown in the drawings is only an example and does not necessarily include all the contents and operations / steps, nor does it necessarily be executed in the described order. For example, some operations / steps can be further divided, combined or partially merged, so that the actual execution order can be changed according to the actual situation.

[0030] The embodiments of the present application can acquire and process related data based on artificial intelligence technology. Artificial intelligence (AI) is the use of digital computers or computer-controlled machines to simulate, extend and expand human intelligence, perceive the environment, acquire knowledge and use knowledge to obtain the best results.

[0031] The basic technology of artificial intelligence generally includes technologies such as sensors, special artificial intelligence chips, cloud computing, distributed storage, big data processing technology, operation / interaction system, mechatronics, etc. The software technology of artificial intelligence mainly includes computer vision technology, robot technology, biometric technology, speech processing technology, natural language processing technology, and machine learning / deep learning, etc.

[0032] With the rapid development of economy, the continuous improvement of national income level and education level, the health insurance of the financial industry has attracted more and more attention, including life insurance, medical insurance, disease insurance and accident insurance, etc. At present, the pricing of health insurance products can only be based on age to make simple policy pricing, and cannot accurately price personalized products for each insured person, resulting in a decline in user recognition of insurance products, and thus high complaint rate and low time competitiveness of insurance products.

[0033] The embodiment of the application provides a kind of based on artificial intelligence's insurance product pricing method, device, equipment and medium, the insurance product pricing method based on artificial intelligence includes obtaining the health data of insured person and health risk assessment model, the health risk assessment model is obtained by training neural network model based on multiple sample data in advance, and the sample data includes sample health data and labeled health risk grade;The health risk grade of the insured person is obtained by carrying out health risk grade evaluation to the health data of the insured person by the health risk assessment model;The base pricing of insurance product is acquired, and the premium floating price corresponding to the insured person is determined according to the health risk grade;According to the base pricing and premium floating price, the target pricing of the insured person to purchase the insurance product is determined.

[0034] Wherein, the insurance product pricing method based on artificial intelligence can be applied in computer equipment, which can be mobile phone, tablet computer, notebook computer, desktop computer, personal digital assistant and wearable device and other electronic devices.For example, the computer equipment can be notebook computer, and notebook computer obtains the health data of insured person and health risk assessment model, the health risk assessment model is obtained by training neural network model based on multiple sample data in advance, and the sample data includes sample health data and labeled health risk grade;The health risk grade of the insured person is obtained by carrying out health risk grade evaluation to the health data of the insured person by the health risk assessment model;The base pricing of insurance product is acquired, and the premium floating price corresponding to the insured person is determined according to the health risk grade;According to the base pricing and premium floating price, the target pricing of the insured person to purchase the insurance product is determined.

[0035] Some embodiments of the application will be described in detail below with reference to the accompanying drawings. The following examples and features in the examples can be combined with each other without conflict.

[0036] Please refer to Figure 1 , Figure 1 A flowchart of the insurance product pricing method based on artificial intelligence provided by the embodiment of the application is shown.

[0037] As Figure 1 The insurance product pricing method based on artificial intelligence includes steps S101-S104.

[0038] Step S101, obtaining the health data of insured person and health risk assessment model.

[0039] Wherein, the health risk assessment model is obtained by training neural network model based on multiple sample data in advance, and the sample data includes sample health data and labeled health risk grade.

[0040] It should be noted that the health data includes health indicators, living habits and medical consultation records, the health indicators are physiological indicators of the human body, and the health indicators include but are not limited to weight, body temperature, heart rate, blood pressure, body fat and the like, the living habits are habits and living rules of people in life, and the living habits include but are not limited to eating habits, exercise habits and sleeping habits and the like; the medical consultation record is a medical record of people inquiring in a medical institution, and the medical consultation record includes but is not limited to cold consultation records and chronic disease consultation records and the like.

[0041] In some embodiments, the computer device establishes a communication connection with an application program for recording health indicators and living habits, and obtains the health indicators and living habits of the insured person from the application program for recording health indicators, which can be collected based on the smart wearable device worn by the insured person, for example, collecting information such as body temperature, heart rate, blood pressure and body fat of the insured person through the smart wearable device. Information such as exercise habits and sleeping habits of the insured person is collected through the smart wearable device. By establishing a communication connection with the application program, the health indicators and living habits of the insured person can be accurately obtained, greatly improving the efficiency and accuracy of health data collection.

[0042] In some embodiments, the computer device establishes a communication connection with a medical system, and obtains the medical consultation records of the insured person from the medical system. The medical system is a system established by various hospitals and clinics, and the medical system records the consultation records of various patients. The medical consultation records of the insured person can be accurately obtained from the medical system, greatly improving the efficiency and accuracy of health data collection.

[0043] In some embodiments, as shown in Figure 2 The method for pricing an insurance product based on artificial intelligence further includes steps S201 to S205.

[0044] Step S201, obtaining a sample data set, the sample data set including a plurality of sample data, the sample data including sample health data and labeled health risk levels.

[0045] The sample data set includes a plurality of sample data, and the sample data includes sample health data and labeled health risk levels.

[0046] In some embodiments, a history data is obtained, the history data including health data and a health risk level that has been set for the health data, the health data is marked as sample health data and the health risk level is marked as a labeled health risk level; the sample health data and the labeled health risk level are taken as a sample data. The foregoing steps are executed in a loop to obtain a sample data set.

[0047] Step S202, obtaining a preset neural network model, and selecting a sample data from the sample data set as a target sample data.

[0048] The neural network model includes, but is not limited to, a convolutional neural network model, a recurrent neural network model, and a recurrent convolutional neural network model, etc. For example, as shown in the figure, Figure 3 The neural network model includes an encoder layer, a full connection layer, and a decoder layer. The encoder layer is used for feature encoding of health data, the full connection layer is used for health risk level prediction, and the decoder layer is used for decoding processing of a feature vector.

[0049] In some embodiments, a sample data is selected from the sample data set as a target sample data, wherein the target sample data includes sample health data and labeled health risk level. By selecting the sample data from the sample data set, the target sample data can be accurately obtained.

[0050] Step S203, performing health risk level evaluation on the sample health data in the target sample data by using the preset neural network model, to obtain a predicted health risk level.

[0051] The sample health data includes sample health indicators, sample living habits, and sample medical interview records.

[0052] In some embodiments, the sample health indicators, the sample living habits, and the sample medical interview records are input into the encoder layer for feature encoding, to obtain a predicted health indicator feature vector, a predicted living habit feature vector, and a predicted medical interview record feature vector; the predicted health indicator feature vector, the predicted living habit feature vector, and the predicted medical interview record feature vector are input into the full connection layer for health risk level prediction, to obtain a predicted health risk level feature vector; and the predicted health risk level feature vector is decoded by the decoder layer, to obtain the predicted health risk level of the insured person.

[0053] In some embodiments, the manner of inputting the predicted health indicator feature vector, the predicted life habit feature vector and the predicted medical consultation record feature vector into the full connection layer for health risk level prediction to obtain the predicted health risk level feature vector can be: performing health risk level prediction on the predicted health indicator feature vector by the full connection layer to obtain a predicted first health risk level feature vector; performing health risk level prediction on the predicted life habit feature vector by the full connection layer to obtain a predicted second health risk level feature vector; performing health risk level prediction on the predicted medical consultation record feature vector by the full connection layer to obtain a predicted third health risk level feature vector; and performing feature fusion on the predicted first health risk level feature vector, the predicted second health risk level feature vector and the predicted third health risk level feature vector to obtain the predicted health risk level feature vector.

[0054] In some embodiments, the manner of performing feature fusion on the predicted first health risk level feature vector, the predicted second health risk level feature vector and the predicted third health risk level feature vector to obtain the predicted health risk level feature vector can be: performing weighting processing on the predicted first health risk level feature vector based on a preset first weight coefficient to obtain a predicted target first health risk level feature vector; performing weighting processing on the predicted second health risk level feature vector based on a preset second weight coefficient to obtain a predicted target second health risk level feature vector; performing weighting processing on the predicted third health risk level feature vector based on a preset third weight coefficient to obtain a predicted target third health risk level feature vector; and performing feature vector fusion on the predicted target first health risk level feature vector, the predicted target second health risk level feature vector and the predicted target third health risk level feature vector to obtain the predicted health risk level feature vector. The preset first weight coefficient, the preset second weight coefficient and the preset third weight coefficient can be set according to actual conditions, and the embodiments of the present application do not make specific limitations thereto.

[0055] Step S204: determining whether the preset neural network model converges according to the labeled health risk level in the target sample data and the predicted health risk level.

[0056] According to the labeled health risk level and the predicted health risk level, a loss value of the preset neural network model is determined; if the loss value is less than or equal to a preset loss value, it is determined that the preset neural network model has converged; if the loss value is greater than the preset loss value, it is determined that the preset neural network model has not converged. The preset loss value is set according to actual conditions, and embodiments of the present application do not make specific limitations thereon, for example, the preset loss value can be set to 0.02. According to the labeled health risk level and the predicted health risk level, the loss value of the preset neural network model can be accurately determined, and whether the neural network model has converged can be accurately known based on the loss value.

[0057] In some embodiments, according to the labeled health risk level and the predicted health risk level, the manner of determining the loss value of the preset neural network model can be: calculating the similarity of the labeled health risk level and the predicted health risk level to obtain a current similarity; obtaining a historical similarity, which is the average of the current similarity of each sample data that has completed training, performing mean calculation on the current similarity and the historical similarity to obtain a target similarity; subtracting the target similarity from 1 to obtain a value, which is determined as the loss value. The manner of calculating the similarity can be selected according to actual conditions, and embodiments of the present application do not make specific limitations thereon, for example, the manner of calculating the similarity can be calculating the cosine similarity of the labeled health risk level and the predicted health risk level.

[0058] Step S205, if the preset neural network model has not converged, adjusting the model parameters of the preset neural network model, and continuing to perform the step of selecting one sample data from the sample data set as a target sample data until a converged health risk assessment model is obtained.

[0059] If the loss value is greater than the preset loss value, it is determined that the preset neural network model has not converged, the model parameters of the preset neural network model are adjusted, and the step of selecting one sample data from the sample data set as a target sample data is continued; the health risk level of the target sample data is evaluated by the preset neural network model to obtain a predicted health risk level; according to the labeled health risk level and the predicted health risk level of the target sample data, the step of determining whether the preset neural network model has converged is performed until a converged health risk assessment model is obtained.

[0060] Step S102, evaluating the health risk level of the health data of the insured person by the health risk assessment model to obtain the health risk level of the insured person.

[0061] The health risk assessment model comprises an encoder layer, a full connection layer, and a decoder layer; the health data comprises health indicators, living habits, and medical interview records. If data in the health indicators, the living habits, and the medical interview records cannot be obtained, empty data packets are used to replace the unobtained data.

[0062] In some embodiments, as shown in FIG. 1, step S102 comprises sub-step S1021 to sub-step S1023. Figure 4

[0063] In some embodiments, the health indicators, the living habits, and the medical interview records are input into the encoder layer for feature encoding to obtain health indicator feature vectors, living habit feature vectors, and medical interview record feature vectors.

[0064] In some embodiments, the health indicators are input into the encoder layer for feature encoding to obtain health indicator feature vectors. The living habits are input into the encoder layer for feature encoding to obtain living habit feature vectors. The medical interview records are input into the encoder layer for feature encoding to obtain medical interview record feature vectors. The health indicators, the living habits, and the medical interview records are input into the encoder layer for feature encoding to accurately obtain the health indicator feature vectors, the living habit feature vectors, and the medical interview record feature vectors, greatly improving the efficiency and accuracy of the insurance product pricing.

[0065] In some embodiments, the health indicator feature vectors, the living habit feature vectors, and the medical interview record feature vectors are input into the full connection layer for health risk level prediction to obtain health risk level feature vectors.

[0066] The health indicator feature vectors are input into the full connection layer for health risk level prediction to obtain first health risk level feature vectors. The living habit feature vectors are input into the full connection layer for health risk level prediction to obtain second health risk level feature vectors. The medical interview record feature vectors are input into the full connection layer for health risk level prediction to obtain third health risk level feature vectors. The first health risk level feature vectors, the second health risk level feature vectors, and the third health risk level feature vectors are fused to obtain health risk level feature vectors. The health indicator feature vectors, the living habit feature vectors, and the medical interview record feature vectors are input into the full connection layer for health risk level prediction to accurately obtain the health risk level feature vectors.

[0067] ​In some embodiments, the manner of performing feature fusion on the first health risk level feature vector, the second health risk level feature vector, and the third health risk level feature vector to obtain the health risk level feature vector can be: performing weighted processing on the first health risk level feature vector based on a preset first weight coefficient to obtain a target first health risk level feature vector; performing weighted processing on the second health risk level feature vector based on a preset second weight coefficient to obtain a target second health risk level feature vector; performing weighted processing on the third health risk level feature vector based on a preset third weight coefficient to obtain a target third health risk level feature vector; and performing feature vector fusion on the target first health risk level feature vector, the target second health risk level feature vector, and the target third health risk level feature vector to obtain the health risk level feature vector. The preset first weight coefficient, the preset second weight coefficient, and the preset third weight coefficient can be set according to actual conditions, and embodiments of the present application do not make specific limitations thereto. For example, the preset first weight coefficient can be set to 0.6, the preset second weight coefficient can be set to 0.3, and the preset third weight coefficient can be set to 0.1.

[0068] Sub-step S1023: performing decoding processing on the health risk level feature vector by the decoder layer to obtain the health risk level of the insured person.

[0069] Performing decoding processing on the health risk level feature vector by the decoder layer can accurately obtain the health risk level of the insured person.

[0070] Step S103: obtaining a basic pricing of an insurance product, and determining a premium floating price corresponding to the insured person according to the health risk level.

[0071] The basic pricing of the insurance product is a fixed basic price formulated by an insurance company for the insurance product. For example, the basic price of a major illness insurance is 500 yuan / year.

[0072] In some embodiments, the basic pricing table of the insurance product formulated by the insurance company is obtained, which can accurately obtain the basic pricing of the insurance product.

[0073] In some embodiments, a mapping relationship table between a preset health risk level and a premium floating price is obtained, and a premium floating price matched with the health risk level is queried from the mapping relationship table to obtain the premium floating price corresponding to the insured person. The mapping relationship table is established in advance based on the health risk level and the premium floating price, and the mapping relationship table can be established according to actual conditions, and embodiments of the present application do not make specific limitations thereto. The mapping relationship table can accurately query the premium floating price corresponding to the insured person.

[0074] An exemplary health risk level of the insured person is three, and a premium float price matched with the health risk level three is queried from the mapping relationship table to obtain a corresponding premium float price of the insured person as 200 yuan.

[0075] In step S104, a target pricing of the insurance product purchased by the insured person is determined according to the basic pricing and the premium float price.

[0076] The basic pricing and the premium float price are added to obtain the target pricing of the insurance product purchased by the insured person. The target pricing of the insurance product can be accurately obtained by superimposing the basic pricing and the premium float price.

[0077] An exemplary health risk level of the insured person is three, and a premium float price matched with the health risk level three is queried from the mapping relationship table to obtain a corresponding premium float price of the insured person as 200 yuan.

[0078] The above embodiment provides an insurance product pricing method based on artificial intelligence. The health data of the insured person and a health risk assessment model are obtained. The health risk assessment model is obtained by pre-training a neural network model based on a plurality of sample data. The sample data includes sample health data and labeled health risk levels. The health risk level of the insured person is obtained by the health risk assessment model. The basic pricing of the insurance product is obtained, and the corresponding premium float price of the insured person is determined according to the health risk level. The target pricing of the insurance product purchased by the insured person is determined according to the basic pricing and the premium float price. In this application, the health risk level of the insured person is accurately obtained by the health risk assessment model, and the corresponding premium float price of the insured person is accurately determined based on the health risk level. The target pricing of the insurance product purchased by the insured person is accurately determined based on the basic pricing and the premium float price of the insurance product. The accuracy of the insurance product pricing is greatly improved, and the satisfaction of the insured person with the insurance product is effectively improved.

[0079] Please refer to Figure 5 , Figure 5 A schematic block diagram of an insurance product pricing device provided by the embodiment of the application is shown.

[0080] As shown in Figure 5 The insurance product pricing apparatus 300 includes an obtaining module 310, a generating module 320, and a determining module 330, wherein:

[0081] The obtaining module 310 is configured to obtain health data of an insured person and a health risk assessment model, the health risk assessment model being obtained by training a neural network model based on a plurality of sample data, the sample data including sample health data and labeled health risk levels;

[0082] The generating module 320 is configured to perform health risk level assessment on the health data of the insured person by using the health risk assessment model, to obtain a health risk level of the insured person;

[0083] The obtaining module 310 is further configured to obtain a basic pricing of an insurance product;

[0084] The determining module 330 is configured to determine a premium floating price corresponding to the insured person according to the health risk level;

[0085] The determining module 330 is further configured to determine a target pricing of the insurance product for the insured person according to the basic pricing and the premium floating price.

[0086] In some embodiments, as shown in Figure 6 The generating module 320 includes a first generating submodule 321, a second generating submodule 322, and a third generating submodule 323, wherein:

[0087] The first generating submodule 321 is configured to input the health indicators, the life habits, and the medical interview records into the encoder layer for feature encoding, to obtain health indicator feature vectors, life habit feature vectors, and medical interview record feature vectors;

[0088] The second generating submodule 322 is configured to input the health indicator feature vectors, the life habit feature vectors, and the medical interview record feature vectors into the fully connected layer for health risk level prediction, to obtain health risk level feature vectors;

[0089] The third generating submodule 323 is configured to perform decoding processing on the health risk level feature vectors by using the decoder layer, to obtain the health risk level of the insured person.

[0090] In some embodiments, the second generating submodule 322 is further configured to:

[0091] perform health risk level prediction on the health indicator feature vectors by using the fully connected layer, to obtain first health risk level feature vectors;

[0092] predicting a health risk level of the life habit feature vector through the full connection layer to obtain a second health risk level feature vector;

[0093] predicting a health risk level of the medical consultation record feature vector through the full connection layer to obtain a third health risk level feature vector;

[0094] performing feature fusion on the first health risk level feature vector, the second health risk level feature vector and the third health risk level feature vector to obtain the health risk level feature vector.

[0095] In some embodiments, the second generation sub-module 322 is further configured to:

[0096] performing weighting processing on the first health risk level feature vector based on a preset first weight coefficient to obtain a target first health risk level feature vector;

[0097] performing weighting processing on the second health risk level feature vector based on a preset second weight coefficient to obtain a target second health risk level feature vector;

[0098] performing weighting processing on the third health risk level feature vector based on a preset third weight coefficient to obtain a target third health risk level feature vector;

[0099] performing feature vector fusion on the target first health risk level feature vector, the target second health risk level feature vector and the target third health risk level feature vector to obtain the health risk level feature vector.

[0100] In some embodiments, the determination module 330 is further configured to:

[0101] obtaining a mapping relationship table between a preset health risk level and a premium floating price;

[0102] querying the premium floating price matched with the health risk level from the mapping relationship table to obtain the premium floating price corresponding to the insured person.

[0103] In some embodiments, the insurance product pricing apparatus 300 is further configured to:

[0104] obtaining a sample data set, the sample data set including a plurality of sample data, the sample data including sample health data and labeled health risk level;

[0105] obtaining a preset neural network model and selecting a sample data from the sample data set as a target sample data;

[0106] The preset neural network model is used for health risk level evaluation on sample health data in the target sample data, to obtain a predicted health risk level;

[0107] According to the labeled health risk level in the target sample data and the predicted health risk level, it is determined whether the preset neural network model converges;

[0108] If the preset neural network model does not converge, the model parameters of the preset neural network model are adjusted, and the process of selecting a sample data from the sample data set as a target sample data is continued until a converged health risk evaluation model is obtained.

[0109] In some embodiments, the insurance product pricing apparatus 300 is further configured to:

[0110] According to the labeled health risk level and the predicted health risk level, a loss value of the preset neural network model is determined;

[0111] If the loss value is less than or equal to a preset loss value, it is determined that the preset neural network model has converged;

[0112] If the loss value is greater than the preset loss value, it is determined that the preset neural network model does not converge.

[0113] It should be noted that, for the convenience and brevity of description, the specific working process of the above-mentioned insurance product pricing apparatus can refer to the corresponding process in the foregoing artificial intelligence-based insurance product pricing method embodiments, which will not be described here.

[0114] Please refer to Figure 7 , Figure 7 A structural schematic block diagram of a computer device according to an embodiment of the present application is provided.

[0115] As Figure 7 shown, the computer device 400 includes a processor 402 and a memory 403 connected through a system bus 401, wherein the memory 403 can include a storage medium and an internal memory.

[0116] The storage medium can store a computer program. The computer program includes program instructions which, when executed, can cause the processor to perform any one of the artificial intelligence-based insurance product pricing methods.

[0117] The processor 402 is configured to provide computing and control capabilities to support the operation of the entire computer device.

[0118] The internal memory provides an environment for the running of a computer program in a storage medium, and the computer program, when executed by the processor, can enable the processor to perform any artificial intelligence-based insurance product pricing method.

[0119] Those skilled in the art can understand that, Figure 7 The skilled in the art can understand that,

[0120] It should be understood that the processor 402 can be a central processing unit (CPU), and the processor can also be other general-purpose processors, digital signal processors (DSP), application specific integrated circuits (ASIC), field programmable gate arrays (FPGA) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. Among them, the general-purpose processor can be a microprocessor or the processor can also be any conventional processor, etc.

[0121] In one embodiment, the processor 402 is configured to run a computer program stored in the memory to perform the following steps:

[0122] Obtain the health data of the insured person and a health risk assessment model, the health risk assessment model being obtained by training a neural network model based on a plurality of sample data, the sample data including sample health data and labeled health risk levels;

[0123] Perform health risk level assessment on the health data of the insured person through the health risk assessment model to obtain the health risk level of the insured person;

[0124] Obtain the basic pricing of the insurance product, and determine the premium floating price corresponding to the insured person according to the health risk level;

[0125] Determine the target pricing of the insurance product purchased by the insured person according to the basic pricing and the premium floating price.

[0126] In an embodiment, the health risk assessment model comprises an encoder layer, a full connection layer, and a decoder layer; the health data comprises health indicators, living habits, and medical interview records; the processor 402, in implementing the health risk level assessment of the health data of the insured person by the health risk assessment model to obtain the health risk level of the insured person, is configured to:

[0127] input the health indicators, the living habits, and the medical interview records into the encoder layer for feature encoding to obtain health indicator feature vectors, living habit feature vectors, and medical interview record feature vectors;

[0128] input the health indicator feature vectors, the living habit feature vectors, and the medical interview record feature vectors into the full connection layer for health risk level prediction to obtain health risk level feature vectors;

[0129] decode the health risk level feature vectors by the decoder layer to obtain the health risk level of the insured person.

[0130] In an embodiment, the processor 402, in implementing the input of the health indicator feature vectors, the living habit feature vectors, and the medical interview record feature vectors into the full connection layer for health risk level prediction to obtain health risk level feature vectors, is configured to:

[0131] predict the health risk level of the health indicator feature vectors by the full connection layer to obtain first health risk level feature vectors;

[0132] predict the health risk level of the living habit feature vectors by the full connection layer to obtain second health risk level feature vectors;

[0133] predict the health risk level of the medical interview record feature vectors by the full connection layer to obtain third health risk level feature vectors;

[0134] fuse the first health risk level feature vectors, the second health risk level feature vectors, and the third health risk level feature vectors to obtain the health risk level feature vectors.

[0135] In an embodiment, the processor 402, in implementing the fusion of the first health risk level feature vectors, the second health risk level feature vectors, and the third health risk level feature vectors to obtain the health risk level feature vectors, is configured to:

[0136] weighting processing on the first health risk level feature vector based on a preset first weight coefficient, to obtain a target first health risk level feature vector;

[0137] weighting processing on the second health risk level feature vector based on a preset second weight coefficient, to obtain a target second health risk level feature vector;

[0138] weighting processing on the third health risk level feature vector based on a preset third weight coefficient, to obtain a target third health risk level feature vector;

[0139] performing feature vector fusion on the target first health risk level feature vector, the target second health risk level feature vector and the target third health risk level feature vector, to obtain the health risk level feature vector.

[0140] In one embodiment, the processor 402, when implementing the determining of the premium float price corresponding to the insured person according to the health risk level, is configured to implement:

[0141] obtain a mapping relationship table between preset health risk levels and premium float prices;

[0142] query the premium float price matched with the health risk level from the mapping relationship table, to obtain the premium float price corresponding to the insured person.

[0143] In one embodiment, the processor 402 is further configured to implement:

[0144] obtain a sample data set, the sample data set including a plurality of sample data, the sample data including sample health data and labeled health risk levels;

[0145] obtain a preset neural network model, and select a sample data from the sample data set as a target sample data;

[0146] perform health risk level evaluation on the sample health data in the target sample data through the preset neural network model, to obtain a predicted health risk level;

[0147] determine whether the preset neural network model converges according to the labeled health risk level in the target sample data and the predicted health risk level;

[0148] if the preset neural network model does not converge, adjust the model parameters of the preset neural network model, and continue to select a sample data from the sample data set as a target sample data until a converged health risk evaluation model is obtained.

[0149] In an embodiment, the processor 402, when implementing the determining whether the preset neural network model converges according to the labeled health risk level and the predicted health risk level in the target sample data, is configured to implement:

[0150] determining a loss value of the preset neural network model according to the labeled health risk level and the predicted health risk level;

[0151] if the loss value is less than or equal to a preset loss value, determining that the preset neural network model has converged;

[0152] if the loss value is greater than the preset loss value, determining that the preset neural network model has not converged.

[0153] It should be noted that, for the convenience and brevity of description, the specific working process of the computer device described above can be clearly understood by those skilled in the art, and the corresponding process in the foregoing embodiments of the insurance product pricing method based on artificial intelligence can be referred to, which will not be described here.

[0154] The embodiments of the present application also provide a computer readable storage medium, the computer readable storage medium stores a computer program, the computer program includes program instructions, and the method implemented by the program instructions when executed can refer to each embodiment of the insurance product pricing method based on artificial intelligence of the present application.

[0155] The computer readable storage medium can be an internal storage unit of the computer device, such as a hard disk or a memory of the computer device. The computer readable storage medium can be non-volatile or volatile. The computer readable storage medium can also be an external storage device of the computer device, such as a plug-in hard disk, a smart media card (SMC), a secure digital (SD) card, a flash card, etc.

[0156] Further, the computer readable storage medium can mainly include a program storage area and a data storage area, wherein the program storage area can store an operating system, at least one application required by a function, etc.; and the data storage area can store data created according to the use of the blockchain node, etc.

[0157] The blockchain referred to in the present application is a new application mode of computer technologies such as distributed data storage, peer-to-peer transmission, consensus mechanism, and encryption algorithm. The blockchain is essentially a decentralized database and is a series of data blocks associated using cryptographic methods. Each data block contains information about a batch of network transactions and is used to verify the validity (anti-fake) of the information and generate the next block. The blockchain can include a blockchain underlying platform, a platform product service layer, and an application service layer.

[0158] It should be understood that the terms used in this application specification are only for the purpose of describing specific embodiments and are not intended to limit the application. As used in this application specification, unless the context clearly indicates otherwise, the singular forms "a", "an", and "the" are intended to include the plural forms.

[0159] It should also be understood that the term "and / or" used in the present application specification means any combination of one or more of the associated listed items and all possible combinations, and includes these combinations. It should be noted that in this document, the terms "comprise", "include" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or system including a series of elements includes not only those elements, but also other elements not explicitly listed, or other elements inherent to such a process, method, article or system. Without more limitations, the element defined by the statement "comprises a" does not exclude the presence of other identical elements in the process, method, article or system including the element.

[0160] The above-mentioned application embodiment serial numbers are only for description, not representing the advantages and disadvantages of the embodiments. The above description is only a specific implementation of the present application, but the protection scope of the present application is not limited thereto. Any person skilled in the art can easily think of various equivalent modifications or replacements within the technical scope disclosed by the present application, and these modifications or replacements should be covered within the protection scope of the present application.

Claims

1. A pricing method for insurance products based on artificial intelligence, characterized in that, include: Obtain the insured's health data and health risk assessment model. The health risk assessment model is obtained by pre-training a neural network model based on multiple sample data. The sample data includes sample health data and labeled health risk levels. The health risk level of the insured is obtained by assessing the health risk level of the insured through the health risk assessment model. Obtain the basic pricing of the insurance product and determine the premium floating price corresponding to the insured based on the stated health risk level; The target price for the insured to purchase the insurance product is determined based on the base price and the premium floating price.

2. The insurance product pricing method based on artificial intelligence as described in claim 1, characterized in that, The health risk assessment model includes an encoder layer, a fully connected layer, and a decoder layer; the health data includes health indicators, lifestyle habits, and medical consultation records. The process of assessing the health risk level of the insured through the health risk assessment model to obtain the insured's health risk level includes: The health indicators, lifestyle habits, and medical consultation records are input into the encoder layer for feature encoding to obtain health indicator feature vectors, lifestyle habit feature vectors, and medical consultation record feature vectors. The health indicator feature vector, the lifestyle habit feature vector, and the medical consultation record feature vector are input into the fully connected layer to predict the health risk level, thereby obtaining the health risk level feature vector. The health risk level feature vector is decoded by the decoder layer to obtain the health risk level of the insured.

3. The artificial intelligence-based insurance product pricing method as described in claim 2, characterized in that, The step of inputting the health indicator feature vector, the lifestyle habit feature vector, and the medical consultation record feature vector into the fully connected layer to predict the health risk level, and obtaining the health risk level feature vector, includes: The health risk level is predicted by the health indicator feature vector through the fully connected layer to obtain the first health risk level feature vector. The health risk level of the lifestyle feature vector is predicted by the fully connected layer to obtain the second health risk level feature vector. The health risk level is predicted by the feature vector of the medical consultation record through the fully connected layer to obtain the third health risk level feature vector. The first health risk level feature vector, the second health risk level feature vector, and the third health risk level feature vector are fused to obtain the health risk level feature vector.

4. The artificial intelligence-based insurance product pricing method as described in claim 3, characterized in that, The step of fusing the first health risk level feature vector, the second health risk level feature vector, and the third health risk level feature vector to obtain the health risk level feature vector includes: The first health risk level feature vector is weighted based on a preset first weight coefficient to obtain the target first health risk level feature vector. The second health risk level feature vector is weighted based on a preset second weight coefficient to obtain the target second health risk level feature vector. The third health risk level feature vector is weighted based on a preset third weighting coefficient to obtain the target third health risk level feature vector. The feature vectors of the first, second, and third health risk levels of the target are fused to obtain the health risk level feature vector.

5. The artificial intelligence-based insurance product pricing method as described in claim 1, characterized in that, The process of determining the premium floating price corresponding to the insured based on the health risk level includes: Obtain the pre-defined mapping table between health risk levels and premium fluctuation prices; The premium floating price corresponding to the health risk level is obtained by querying the mapping table.

6. The insurance product pricing method based on artificial intelligence as described in claim 1, characterized in that, The method further includes: Obtain a sample dataset, which includes multiple sample data, including sample health data and labeled health risk levels; Obtain a preset neural network model, and select a sample data from the sample dataset as the target sample data; The health risk level of the sample health data in the target sample data is assessed by the preset neural network model to obtain the predicted health risk level. Based on the labeled health risk level in the target sample data and the predicted health risk level, determine whether the preset neural network model has converged; If the preset neural network model fails to converge, the model parameters of the preset neural network model are adjusted, and the process of selecting a sample data from the sample dataset as the target sample data continues until a converged health risk assessment model is obtained.

7. The artificial intelligence-based insurance product pricing method as described in claim 6, characterized in that, The step of determining whether the preset neural network model has converged based on the labeled health risk level in the target sample data and the predicted health risk level includes: Based on the labeled health risk level and the predicted health risk level, determine the loss value of the preset neural network model; If the loss value is less than or equal to the preset loss value, it is determined that the preset neural network model has converged; If the loss value is greater than the preset loss value, it is determined that the preset neural network model has not converged.

8. An insurance product pricing device, characterized in that, The insurance product pricing device includes an acquisition module, a generation module, and a determination module, wherein: The acquisition module is used to acquire the insured's health data and health risk assessment model. The health risk assessment model is obtained by training a neural network model in advance based on multiple sample data. The sample data includes sample health data and labeled health risk levels. The generation module is used to assess the health risk level of the insured's health data through the health risk assessment model, and obtain the health risk level of the insured. The acquisition module is also used to acquire the basic pricing of insurance products; The determining module is used to determine the premium floating price corresponding to the insured based on the health risk level; The determining module is further configured to determine the target price for the insured to purchase the insurance product based on the base price and the premium floating price.

9. A computer device, characterized in that, The computer device includes a processor, a memory, and a computer program stored in the memory and executable by the processor, wherein when the computer program is executed by the processor, it implements the steps of the artificial intelligence-based insurance product pricing method as described in any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a computer program, wherein when the computer program is executed by a processor, it implements the steps of the artificial intelligence-based insurance product pricing method as described in any one of claims 1 to 7.

Images