Insurance policy renewal rate estimation method and device, electronic equipment, medium and product
By using the payment prediction model in the insurance industry to estimate the payment probability of the policy, and combining the policy period premium, the actual payment total and the target period premium, the high-cost problem of estimating the policy renewal rate in the existing technology is solved. problem, achieving efficient and flexible renewal rate estimation.
Patent Information
- Application Number
- CN202411327179.1
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-09-23
- Publication Date
- 2025-01-21
AI Technical Summary
In the prior art, statistical methods are used to estimate the renewal rate of insurance policies in specific aggregate dimensions such as time, product and other dimensions, which requires high labor and time costs.
By obtaining the insurance policy data of the target policy, the payment of the payment estimation model is obtained, and the payment probability is obtained, and the policy of policy policy is determined according to the payment probability, the policy premium period, the actual payment of the actual payment and the target period.
It reduces labor and time costs, enables flexible estimation of policy renewal rates, and improves estimation efficiency.
Smart Images
Figure CN119338597A_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the technical field of renewal rate estimation, and in particular to a method, device, electronic equipment, medium and product for estimating policy renewal rate. Background Art
[0002] In recent years, with the rise of Internet insurance platforms, short-term insurance products, mainly medical and accidental insurance, have occupied an increasing share. Compared with traditional insurance sales channels, Internet insurance platforms can obtain more user-side behavioral data, and thus use big data, machine learning and other technologies to estimate the renewal behavior of insured users.
[0003] The renewal behavior of insured users is related to premium income. It is very important to estimate the renewal rate of short-term insurance policies at various stages of protection. At present, in the existing technology, insurance actuaries, analysts and other practitioners can use statistical methods to estimate the renewal rate of insurance policies in specific aggregation dimensions such as time and product based on historical insurance policy renewal data. However, using human factors to estimate the renewal rate of insurance policies from specific aggregation dimensions such as time and product requires high manpower and time costs.
[0004] Therefore, there is an urgent need for a method for estimating the policy renewal rate in order to flexibly estimate the policy renewal rate. Summary of the invention
[0005] The present invention provides a method, device, electronic device, medium and product for estimating policy renewal rate, so as to solve the defect of estimating policy renewal rate in specific aggregation dimensions, such as time, product and other dimensions, using statistical methods in the prior art, and to estimate the payment probability of the policy based on a payment estimation model, and further determine the policy renewal rate of the target policy through the estimated payment probability, policy period premium, actual total payment and target period premium, thereby reducing manpower and time costs and achieving the effect of flexibly estimating the policy renewal rate of the policy.
[0006] The present invention provides a method for estimating a policy renewal rate, comprising the following steps.
[0007] Get the policy data of the target policy.
[0008] The insurance policy data is input into the payment estimation model to obtain the payment probability output by the payment estimation model; wherein the payment estimation model is trained based on the insurance policy sample data, and the insurance policy sample data is the sample data obtained in advance.
[0009] The estimated total payment amount is determined based on the payment probability and the policy period premium; wherein, the policy period premium refers to the premium corresponding to all periods of the target policy; the estimated total payment amount refers to the sum of the payment amounts corresponding to the periods of the target policy that have not yet reached the payment time.
[0010] The total number of payments per installment is determined based on the estimated total number of payments and the actual total number of payments; among which, the actual total number of payments refers to the sum of payment amounts corresponding to the installments of the target policy that have reached the payment time, and the total number of payments per installment refers to the sum of actual payments made for all installments of the target policy.
[0011] The policy renewal rate of the target policy is determined based on the total number of installment payments and the target installment premium; where the target installment premium refers to the sum of all installment payables of the target policy.
[0012] According to a method for estimating a policy renewal rate provided by the present invention, policy data is input into a payment estimation model to obtain a payment probability output by the payment estimation model, including: inputting the policy data into a data layer of the payment estimation model to obtain policy feature data output by the data layer; wherein the data layer is used to perform data docking and data feature processing on the policy data; inputting the policy feature data into a model layer of the payment estimation model to obtain a payment probability output by the model layer; wherein the model layer is used to perform payment prediction on the policy feature data.
[0013] According to a method for estimating a policy renewal rate provided by the present invention, a payment estimation model is trained based on the following steps: Obtaining sample policy data; wherein the sample policy data includes historical payment behavior data, sample data of paying users and non-paying users; inputting the historical payment behavior data, sample data of paying users and non-paying users into the basic classification model to obtain model parameters output by the basic classification model; wherein the basic classification model is a predetermined classification model; optimizing the basic classification model according to the model parameters to obtain a payment estimation model.
[0014] According to a method for estimating policy renewal rate provided by the present invention, the payment probability includes the payment probability corresponding to each period of the target policy; the policy period premium includes the period premium corresponding to each period of the target policy; the estimated total payment is determined based on the payment probability and the policy period premium, including: determining the estimated period payment for each period corresponding to the target policy based on the product of the period payment probability and the period premium for each period corresponding to the target policy; and determining the estimated total payment based on the sum of the estimated period payments for each period.
[0015] According to a method for estimating policy renewal rate provided by the present invention, the total number of payments per period is determined based on the estimated total number of payments and the actual total number of payments, including: obtaining the payment results for each period of the target policy; determining the actual total number of payments based on the payment results for each period; and determining the total number of payments per period based on the sum of the estimated total number of payments and the actual total number of payments.
[0016] According to a method for estimating a policy renewal rate provided by the present invention, after determining the policy renewal rate of a target policy based on the total number of payments per period and the target premium per period, the method further includes: obtaining a target query task; determining a query result corresponding to the target query task from the policy renewal rate based on the target query task; wherein the query result is used to represent the renewal rate of the dimension corresponding to the target query task in the policy renewal rate.
[0017] The present invention also provides a device for estimating policy renewal rate, comprising the following modules.
[0018] The policy data acquisition module is used to obtain the policy data of the target policy.
[0019] The payment probability output module is used to input the insurance policy data into the payment estimation model to obtain the payment probability output by the payment estimation model; wherein the payment estimation model is trained based on the insurance policy sample data, and the insurance policy sample data is the sample data obtained in advance.
[0020] The estimated payment determination module is used to determine the estimated total payment based on the payment probability and the policy period premium; wherein the policy period premium refers to the premium corresponding to all periods of the target policy; the estimated total payment refers to the sum of the payment amounts corresponding to the periods of the target policy that have not yet reached the payment time.
[0021] The installment payment determination module is used to determine the total installment payment based on the estimated total payment and the actual total payment; wherein the actual total payment refers to the sum of payment amounts corresponding to the installments of the target policy that have reached the payment time, and the total installment payment refers to the sum of actual payment amounts for all installments of the target policy.
[0022] The renewal rate determination module determines the policy renewal rate of the target policy based on the total number of payments per period and the target premium per period; wherein the target premium per period refers to the sum of all payable fees for the target policy per period.
[0023] The present invention also provides an electronic device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, it implements any of the above-mentioned methods for estimating the policy renewal rate.
[0024] The present invention also provides a non-transitory computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements any of the above-mentioned methods for estimating the policy renewal rate.
[0025] The present invention also provides a computer program product, including a computer program, which, when executed by a processor, implements any of the above-mentioned methods for estimating the policy renewal rate.
[0026] The present invention provides a method, device, electronic device, medium and product for estimating policy renewal rate, which obtains policy data of a target policy; inputs the policy data into a payment estimation model to obtain a payment probability output by the payment estimation model; wherein the payment estimation model is trained based on policy sample data, and the policy sample data is pre-acquired sample data; an estimated total payment amount is determined based on the payment probability and the policy premium per period; wherein the policy premium per period refers to the premium corresponding to all periods of the target policy; the estimated total payment amount refers to the sum of payment amounts corresponding to periods of the target policy that have not reached the payment time; the total payment amount per period is determined based on the estimated total payment amount and the actual total payment amount; wherein the actual total payment amount refers to the sum of payment amounts corresponding to periods of the target policy that have reached the payment time, and the total payment amount per period refers to the sum of actual paid expenses for all periods of the target policy; the policy renewal rate of the target policy is determined based on the total payment amount per period and the target premium per period; wherein the target premium per period refers to the sum of payable expenses for all periods of the target policy. The technical solution of the present invention is used to solve the defects of the prior art in estimating the renewal rate of insurance policies in specific aggregation dimensions, such as time, product, etc., by using statistical methods, and to estimate the payment probability of insurance policies based on a payment estimation model, and further determine the policy renewal rate of the target policy through the estimated payment probability, policy period premium, actual total payment and target period premium, thereby reducing manpower and time costs and achieving the effect of flexibly estimating the policy renewal rate of insurance policies. BRIEF DESCRIPTION OF THE DRAWINGS
[0027] In order to more clearly illustrate the technical solutions in the present invention or the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.
[0028] Figure 1 It is a flow chart of the method for estimating the policy renewal rate provided by the present invention.
[0029] Figure 2 It is a schematic diagram of querying the policy renewal rate provided by the present invention.
[0030] Figure 3 It is a structural schematic diagram of the device for estimating the policy renewal rate provided by the present invention.
[0031] Figure 4 It is a structural schematic diagram of the electronic device provided by the present invention. DETAILED DESCRIPTION
[0032] In order to make the purpose, technical solution and advantages of the present invention clearer, the technical solution of the present invention will be clearly and completely described below in conjunction with the drawings of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.
[0033] Combine the following Figure 1 The present invention describes a method for estimating a policy renewal rate. The method for estimating a policy renewal rate provided by the present invention can be applicable to estimating insurance policy renewal rates. The executor of the method can be an electronic device or a device for estimating a policy renewal rate disposed in the electronic device. The device for estimating a policy renewal rate can be implemented by software, hardware or a combination of both. Figure 1 is a flow chart of the method for estimating the policy renewal rate provided by the present invention, such as Figure 1 As shown, the method includes: step 101, step 102, step 103, step 104 and step 105.
[0034] Step 101: Obtain policy data of a target policy.
[0035] In this step, the target policy refers to the policy for which the policy renewal rate needs to be estimated, and the policy renewal rate refers to the ratio used for statistical analysis of the payment of the target policy; the policy data is used to reflect the basic information of the target policy, and the basic information may include, for example, the policy premium, the policy coverage amount, the policy order time, the current status of the policy, etc. This embodiment does not limit this.
[0036] The current status of the insurance policy includes whether the insurance policy is currently in place, the periods for which premiums have been paid, the number of days insured for the current period of the insurance policy, etc. This embodiment does not limit this.
[0037] Specifically, the target insurance policy is determined from the data table of the big data, and the policy information of the target insurance policy is obtained, such as the insurance premium, insurance coverage amount, policy order time, current status of the insurance policy, etc. The data table of the big data stores multiple insurance policies and the insurance policy data corresponding to the multiple insurance policies.
[0038] Step 102: input the policy data into the payment estimation model to obtain the payment probability output by the payment estimation model.
[0039] In this step, the payment estimation model is trained based on the insurance policy sample data, and the insurance policy sample data is pre-acquired sample data. The sample data includes pre-acquired policy premium samples, policy guarantee amount samples, policy order time samples, and policy current status samples, etc. This embodiment does not limit this.
[0040] Specifically, after obtaining the insurance policy data, the insurance policy data is input into the payment estimation model to obtain the payment probability output by the payment estimation model.
[0041] In a specific implementation, the specific implementation method of inputting the policy data into the payment estimation model and obtaining the payment probability output by the payment estimation model includes: inputting the policy data into the data layer of the payment estimation model to obtain the policy characteristic data output by the data layer; wherein the data layer is used to perform data docking and data feature processing on the policy data; inputting the policy characteristic data into the model layer of the payment estimation model to obtain the payment probability output by the model layer; wherein the model layer is used to perform payment prediction on the policy characteristic data.
[0042] In this step, the data layer performs data docking and data feature processing on the insurance policy data. Specifically, different features of different insurance policy data are processed differently to prepare for the payment estimation model.
[0043] Exemplarily, numerical features such as age are segmented and converted into categorical features. For example, age is divided into groups of 5 years old, such as 0-5 years old, 6-10 years old, 11-15 years old, ..., 95-100 years old, 100 years old +, etc., to represent users of different age groups; features such as gender, province, and city are directly used as categorical features to represent the basic attributes of users; numerical features such as the policy period and the number of days in the current period are retained as status features of the policy; the user's behavior during the insurance process is processed into behavioral features, such as the time from entering the page to placing the order, the duration of each page operation, and other behavioral features; the final output features include: static attributes of the user such as age, gender, province, and city; behavioral features of the user's feedback during the insurance process; characteristics of the time dimension such as the period and the number of days in the current period.
[0044] The model layer makes payment predictions based on the policy feature data. It mainly uses the policy feature data processed by the data layer to train the payment estimation model and estimate the payment probability of each policy.
[0045] Specifically, the policy data is input into the data layer of the payment estimation model, and different categories of features are extracted from the input policy data through the data layer, so as to determine the output policy feature data, such as the user's static attributes such as age, gender, province, and city; the behavioral characteristics of the user's feedback during the insurance process; and the characteristics of the time dimension such as the term, the number of days in the current term, etc. Then, the policy feature data is input into the model layer of the payment estimation model, and the policy feature data is processed and trained through the model layer, so as to determine the payment probability output by the model layer.
[0046] Table 1
[0047] For example, for the payment probability of the target policy, if the target policy is divided into 12 periods, an initial payment probability is estimated for each period, and a total of 12 payment sub-models are trained in the payment estimation model, and the 12 payment sub-models together form a payment estimation model for estimating the renewal rate, and then the payment probabilities corresponding to the 12 periods of the policy are obtained according to the 12 payment sub-models, and the final payment probability of the target policy is determined based on the 12 payment probabilities. As shown in Table 1, Table 1 mainly shows the payment probabilities corresponding to each period of target policies such as Policy 00001 and Policy 00002.
[0048] In a specific embodiment, the payment estimation model is trained based on the following steps: obtaining policy sample data; wherein the policy sample data includes historical payment behavior data, sample data of paying users and non-paying users; inputting the historical payment behavior data, sample data of paying users and non-paying users into a basic classification model to obtain model parameters output by the basic classification model; wherein the basic classification model is a predetermined classification model; optimizing the basic classification model according to the model parameters to obtain a payment estimation model.
[0049] In this step, the policy sample data includes historical payment behavior data, sample data of paying users and non-paying users, and the basic classification model is a predetermined classification model. The basic classification model can be, for example, a gradient boosting decision tree (GBDT), which is an iterative decision tree algorithm.
[0050] Specifically, the historical payment behavior data and paying users in the policy sample data are used as positive sample data, and the sample data of non-paying users are used as negative samples. The positive sample data are input into the basic classification model for model training, and the optimal model parameters are found through the grid search method. The basic classification model is optimized based on the optimal model parameters to obtain the payment estimation model.
[0051] The advantage of this setting is that by training the basic classification model, the payment estimation model can be determined, the accuracy of the payment estimation model can be improved, and the efficiency and comprehensiveness of the payment probability can be improved.
[0052] Step 103: Determine the estimated total payment amount based on the payment probability and the policy period premium.
[0053] In this step, the policy period premium refers to the premium corresponding to all periods of the target policy; the estimated total payment amount refers to the sum of the payment amounts corresponding to the periods of the target policy that have not reached the payment time.
[0054] In a specific embodiment, the specific implementation method of determining the estimated total payment based on the payment probability and the policy period premium includes: determining the estimated payment for each period corresponding to the target policy based on the product of the period payment probability and the period premium for each period corresponding to the target policy; determining the estimated total payment based on the sum of the estimated payment for each period.
[0055] In this step, the payment probability includes the payment probability corresponding to each period of the target policy; the policy period premium includes the period premium corresponding to each period of the target policy.
[0056] Specifically, Wherein, N represents the number of periods that have been paid so far, for example, it can be 1, 2, 12, etc., which is not limited in this embodiment.
[0057] The advantage of this setting is that by determining the payment probability corresponding to each period of the target policy, and then determining the final estimated total payment based on the sum of the products of the payment probability corresponding to each period and the period premium corresponding to each period, the accuracy of the estimated total payment is improved.
[0058] Step 104: Determine the total payment amount for each session based on the estimated total payment amount and the actual total payment amount.
[0059] In this step, the actual total payment refers to the sum of payment amounts corresponding to the periods of the target policy that have reached the payment time, and the total payment per period refers to the sum of actual payment amounts for all periods of the target policy.
[0060] Exemplarily, the actual total payment may be the sum of the actual payment for 1 period, the actual payment for 2 periods, and the actual payment for N periods of the target policy, where N represents the periods that have been paid so far.
[0061] In a specific embodiment, the specific implementation method of determining the total number of payments per installment based on the estimated total number of payments and the actual total number of payments includes: obtaining the payment results for each installment of the target insurance policy; determining the actual total number of payments based on the payment results for each installment; and determining the total number of payments per installment based on the sum of the estimated total number of payments and the actual total number of payments.
[0062] In this step, the payment results of each period are used to determine whether each period of the target policy has been paid, as well as the current payment status of each period, such as unpaid, paid, or actual amount paid, etc. This embodiment does not limit this.
[0063] Specifically, obtain the payment results of each period of the target policy, such as unpaid, paid or actual paid amount, and then determine the payment status of each period of the target policy according to the unpaid, paid or actual paid amount, and then determine the actual total payment based on the payment status of each period of the target policy, and then determine the total payment for each period based on the sum of the actual total payment and the estimated total payment.
[0064] Exemplarily, the total number of payments per installment = the actual total number of payments + the estimated total number of payments, the actual total number of payments = the actual amount paid for 1 installment + the actual amount paid for 2 installments + … + the actual amount paid for N installments, N represents the current payment up to N installments, N is a positive integer and N is not equal to zero.
[0065] Step 105: Determine the policy renewal rate of the target policy based on the total number of payment periods and the target premium period.
[0066] In this step, the target premium refers to the sum of all payable expenses for the target policy.
[0067] Specifically, the policy renewal rate of the target policy is determined based on the sum of the total number of payments per period and the target premium per period, that is, policy renewal rate = total number of payments per period / target premium per period.
[0068] In a specific embodiment, after determining the policy renewal rate of the target policy based on the total number of payments per period and the target premium per period, it also includes: obtaining a target query task; determining a query result corresponding to the target query task from the policy renewal rate based on the target query task; wherein the query result is used to represent the renewal rate of the dimension corresponding to the target query task in the policy renewal rate.
[0069] In this step, the target query task is a task that requires a policy query, and the target query task includes the policy type to be queried, basic policy information, etc., which are not limited in this embodiment. The target query task can be queried from different dimensions, such as by month, by age group, by gender, by insurance product, by month, by age group, etc., which are not limited in this embodiment.
[0070] Specifically, a target query task is obtained; a query result corresponding to the target query task is determined from the policy renewal rate according to the target query task; wherein the query result is used to represent the renewal rate of the dimension corresponding to the target query task in the policy renewal rate.
[0071] Exemplarily, a target query task is obtained, and the target query task is to query the policy renewal rate of the policies of the user group under 30 years old in January XX, and then determine the policy renewal rate of the policies that meet the user group under 30 years old in January XX from all the policies in the data table of all big data through the target query task, so as to determine the query result. The policy renewal rate in the query result is the policy renewal rate of all policies that meet the user group under 30 years old in January XX = the sum of the estimated total payment of the policies that meet the user group under 30 years old in January XX / the sum of the policy premiums of all policies that meet the user group under 30 years old in January XX, as shown in Table 2, which includes the insurance month of the policy, the number of insurance policies underwritten, and the corresponding policy renewal rate of the policy.
[0072] Table 2
[0073] The advantage of this setting is that the corresponding policy renewal rate can be determined in a timely manner based on the target query task, thereby improving the query speed and reducing labor costs.
[0074] For example, the policy renewal rate can also be queried from different dimensions. Figure 2 Schematic diagram of querying the policy renewal rate provided by the present invention, such as Figure 2 As shown in the figure, it shows the distribution diagram of XXX rate (policy renewal rate) by age group, that is, the policy renewal rate is queried from the age group dimension. The age groups are divided into under 30 years old, 31-40 years old, 41-50 years old, 51-60 years old, and over 60 years old. The corresponding policy renewal rates for different age groups are as follows Figure 2 shown.
[0075] The advantage of this setting is that by directly querying the policy renewal rates in different dimensions, the manpower and time costs of querying the policy renewal rates can be saved.
[0076] The present invention provides a method for estimating a policy renewal rate, which comprises the following steps: obtaining policy data of a target policy; inputting the policy data into a payment estimation model to obtain a payment probability output by the payment estimation model; wherein the payment estimation model is trained based on policy sample data, and the policy sample data is sample data acquired in advance; determining an estimated total payment amount according to the payment probability and the policy premium per period; wherein the policy premium per period refers to the premium corresponding to all periods of the target policy respectively; the estimated total payment amount refers to the sum of payment amounts corresponding to periods of the target policy that have not reached the payment time; determining the total payment amount per period according to the estimated total payment amount and the actual total payment amount; wherein the actual total payment amount refers to the sum of payment amounts corresponding to periods of the target policy that have reached the payment time, and the total payment amount per period refers to the sum of actual paid expenses for all periods of the target policy; determining the policy renewal rate of the target policy according to the total payment amount per period and the target premium per period; wherein the target premium per period refers to the sum of payable expenses for all periods of the target policy. On the basis of the above-mentioned embodiments, the technical solution of the present invention is used to solve the defects of the prior art in estimating the renewal rate of insurance policies in specific aggregation dimensions, such as time, product, etc., by using statistical methods, and to estimate the payment probability of insurance policies based on payment estimation models. The renewal rate of the target policy is further determined by the estimated payment probability, policy period premium, actual total payment and target period premium, thereby reducing manpower and time costs and achieving the effect of flexibly estimating the policy renewal rate of insurance policies.
[0077] The following is a description of the device for estimating the policy renewal rate provided by the present invention. The device for estimating the policy renewal rate described below and the method for estimating the policy renewal rate described above can be referenced to each other.
[0078] Figure 3 is a schematic diagram of the structure of the device for estimating the policy renewal rate provided by the present invention, referring to Figure 3 As shown, the device 300 for estimating the policy renewal rate includes: a policy data acquisition module 301, a payment probability output module 302, an estimated payment determination module 303, an installment payment determination module 304 and a renewal rate determination module 305.
[0079] The policy data acquisition module 301 is used to acquire the policy data of the target policy.
[0080] The payment probability output module 302 is used to input the insurance policy data into the payment estimation model to obtain the payment probability output by the payment estimation model; wherein the payment estimation model is trained based on the insurance policy sample data, and the insurance policy sample data is the sample data obtained in advance.
[0081] The estimated payment determination module 303 is used to determine the estimated total payment based on the payment probability and the policy period premium; wherein the policy period premium refers to the premium corresponding to all periods of the target policy; the estimated total payment refers to the sum of the payment amounts corresponding to the periods of the target policy that have not reached the payment time.
[0082] The installment payment determination module 304 is used to determine the total number of installment payments based on the estimated total number of payments and the actual total number of payments; wherein the actual total number of payments refers to the sum of payment amounts corresponding to the installments of the target policy that have reached the payment time, and the total number of installment payments refers to the sum of actual payments made for all installments of the target policy.
[0083] The renewal rate determination module 305 determines the policy renewal rate of the target policy based on the total number of payments per period and the target premium per period; wherein the target premium per period refers to the sum of all payable fees for the target policy per period.
[0084] In an exemplary embodiment, the payment probability output module 302 is specifically used to: input the policy data into the data layer of the payment estimation model to obtain the policy feature data output by the data layer; wherein the data layer is used to perform data docking and data feature processing on the policy data; input the policy feature data into the model layer of the payment estimation model to obtain the payment probability output by the model layer; wherein the model layer is used to perform payment prediction on the policy feature data.
[0085] In an exemplary embodiment, the payment probability output module 302, the training of the payment estimation model, is specifically used for: obtaining policy sample data; wherein the policy sample data includes historical payment behavior data, sample data of paying users and non-paying users; inputting the historical payment behavior data, sample data of paying users and non-paying users into the basic classification model to obtain model parameters output by the basic classification model; wherein the basic classification model is a predetermined classification model; optimizing the basic classification model according to the model parameters to obtain the payment estimation model.
[0086] In an exemplary embodiment, the estimated payment determination module 303 is specifically used to: determine the estimated payment for each period corresponding to the target policy based on the product of the period payment probability and the period premium for each period corresponding to the target policy; determine the total estimated payment based on the sum of the estimated payment for each period.
[0087] In an exemplary embodiment, the installment payment determination module 304 is specifically used to: obtain the installment payment results of the target insurance policy; determine the actual total payment amount based on the installment payment results; and determine the total installment payment amount based on the sum of the estimated total payment amount and the actual total payment amount.
[0088] In an example embodiment, the device further includes: a result query module. The result query module is used to: after determining the policy renewal rate of the target policy according to the total number of payment periods and the target premium period, obtain a target query task; determine a query result corresponding to the target query task from the policy renewal rate according to the target query task; wherein the query result is used to represent the renewal rate of the dimension corresponding to the target query task in the policy renewal rate.
[0089] The device of this embodiment can be used to execute the method of any embodiment in the embodiment of the method for estimating the policy renewal rate. Its specific implementation process and technical effects are similar to those in the embodiment of the method for estimating the policy renewal rate. For details, please refer to the detailed description in the embodiment of the method for estimating the policy renewal rate, which will not be repeated here.
[0090] Figure 4 is a schematic diagram of the structure of the electronic device provided by the present invention, such as Figure 4 As shown, the electronic device may include: a processor (processor) 410 , a communication interface (Communications Interface) 420 , a memory (memory) 430 and a communication bus 340 , wherein the processor 410 , the communication interface 420 , and the memory 430 communicate with each other through the communication bus 340 . The processor 410 can call the logic instructions in the memory 430 to execute the method for estimating the policy renewal rate, which method includes: obtaining the policy data of the target policy; inputting the policy data into the payment estimation model to obtain the payment probability output by the payment estimation model; wherein the payment estimation model is trained based on the policy sample data, and the policy sample data is the sample data obtained in advance; determining the estimated total payment amount according to the payment probability and the policy period premium; wherein the policy period premium refers to the premium corresponding to all periods of the target policy; the estimated total payment amount refers to the sum of the payment amounts corresponding to the periods of the target policy that have not reached the payment time; determining the total payment amount for each period according to the estimated total payment amount and the actual payment amount; wherein the actual payment amount refers to the sum of the payment amounts corresponding to the periods of the target policy that have reached the payment time, and the total payment amount for each period refers to the sum of the actual paid expenses for all periods of the target policy; determining the policy renewal rate of the target policy according to the total payment amount for each period and the target period premium; wherein the target period premium refers to the sum of the payable expenses for all periods of the target policy.
[0091] In addition, the logic instructions in the above-mentioned memory 430 can be implemented in the form of a software functional unit and can be stored in a computer-readable storage medium when it is sold or used as an independent product. Based on such an understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art or the part of the technical solution, can be embodied in the form of a software product, and the computer software product is stored in a storage medium, including a number of instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to perform all or part of the steps of the methods of various embodiments of the present invention. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), disk or optical disk, etc. Various media that can store program codes.
[0092] On the other hand, the present invention also provides a computer program product, which includes a computer program. The computer program can be stored on a non-transitory computer-readable storage medium. When the computer program is executed by a processor, the computer can execute the policy renewal rate estimation method provided by the above methods, which includes: obtaining policy data of the target policy; inputting the policy data into a payment estimation model to obtain the payment probability output by the payment estimation model; wherein the payment estimation model is obtained by training based on policy sample data, and the policy sample data is pre-acquired sample data; determining the estimated payment according to the payment probability and the policy period premium; Total premium; among which, policy period premium refers to the premium corresponding to all periods of the target policy; estimated total payment refers to the sum of payment amounts corresponding to periods of the target policy that have not yet reached the payment time; determine the total period payment based on the estimated total payment and the actual total payment; among which, the actual total payment refers to the sum of payment amounts corresponding to periods of the target policy that have reached the payment time, and the total period payment refers to the sum of actual paid expenses for all periods of the target policy; determine the policy renewal rate of the target policy based on the total period payment and the target period premium; among which, the target period premium refers to the sum of payable expenses for all periods of the target policy.
[0093] On the other hand, the present invention also provides a non-transitory computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements a method for estimating the policy renewal rate provided by the above-mentioned methods, the method comprising: obtaining policy data of a target policy; inputting the policy data into a payment estimation model to obtain a payment probability output by the payment estimation model; wherein the payment estimation model is obtained by training based on policy sample data, and the policy sample data is sample data acquired in advance; determining an estimated total payment based on the payment probability and the policy period premium; wherein the policy period premium Premium refers to the premium corresponding to all periods of the target policy; the estimated total payment refers to the sum of the payment amounts corresponding to the periods of the target policy that have not yet reached the payment time; the total payment per period is determined based on the estimated total payment and the actual total payment; among which, the actual total payment refers to the sum of the payment amounts corresponding to the periods of the target policy that have reached the payment time, and the total payment per period refers to the sum of the actual paid expenses for all periods of the target policy; the policy renewal rate of the target policy is determined based on the total payment per period and the target premium per period; among which, the target premium per period refers to the sum of the payable expenses for all periods of the target policy.
[0094] The device embodiments described above are merely illustrative, wherein the units described as separate components may or may not be physically separated, and the components displayed as units may or may not be physical units, that is, they may be located in one place, or they may be distributed on multiple network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the scheme of this embodiment. Ordinary technicians in this field can understand and implement it without paying creative labor.
[0095] Through the description of the above implementation methods, those skilled in the art can clearly understand that each implementation method can be implemented by means of software plus a necessary general hardware platform, and of course, can also be implemented by hardware. Based on this understanding, the above technical solution is essentially or the part that contributes to the prior art can be embodied in the form of a software product, and the computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, a disk, an optical disk, etc., including a number of instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to execute the methods described in each embodiment or some parts of the embodiments.
[0096] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present invention.
Claims
1. A method for estimating policy renewal rate, characterized in that: include: Obtain policy data of the target policy; Inputting the policy data into a payment estimation model to obtain a payment probability output by the payment estimation model; wherein the payment estimation model is trained based on policy sample data, and the policy sample data is pre-acquired sample data; Determine the estimated total payment amount according to the payment probability and the policy period premium; wherein the policy period premium refers to the premium corresponding to all periods of the target policy; the estimated total payment amount refers to the sum of the payment amounts corresponding to the periods of the target policy that have not reached the payment time; Determine the total number of installments based on the estimated total number of payments and the actual total number of payments; wherein the actual total number of payments refers to the sum of payment amounts corresponding to installments of the target policy that have reached the payment time, and the total number of installments refers to the sum of actual payment amounts for all installments of the target policy; The policy renewal rate of the target policy is determined based on the total number of payments made during each period and the target premium during each period; wherein the target premium during each period refers to the sum of all payable fees for all periods of the target policy.
2. The method for estimating the policy renewal rate according to claim 1, characterized in that: The step of inputting the policy data into a payment estimation model to obtain a payment probability output by the payment estimation model includes: Inputting the policy data into the data layer of the payment estimation model to obtain the policy feature data output by the data layer; wherein the data layer is used to perform data docking and data feature processing on the policy data; The policy characteristic data is input into the model layer of the payment estimation model to obtain the payment probability output by the model layer; wherein the model layer is used to make payment predictions for the policy characteristic data.
3. The method for estimating the policy renewal rate according to claim 2, characterized in that: The payment estimation model is trained based on the following steps: Acquire the policy sample data; wherein the policy sample data includes historical payment behavior data, sample data of paying users and non-paying users; Inputting the historical payment behavior data, the sample data of the paying users and the non-paying users into a basic classification model to obtain model parameters output by the basic classification model; wherein the basic classification model is a predetermined classification model; The basic classification model is optimized according to the model parameters to obtain the payment estimation model.
4. The method for estimating the policy renewal rate according to claim 1, characterized in that: The payment probability includes the payment probability corresponding to each term of the target policy; the policy term premium includes the term premium corresponding to each term of the target policy; and the estimated total payment amount is determined based on the payment probability and the policy term premium, including: Determining the estimated payment for each period corresponding to the target policy according to the product of the payment probability for each period corresponding to the target policy and the premium for each period; The total estimated payment amount is determined based on the sum of the estimated payment amounts for each period.
5. The method for estimating the policy renewal rate according to claim 1, characterized in that: Determining the total payment amount per session according to the estimated total payment amount and the actual total payment amount includes: Obtain the payment results of each period of the target insurance policy; Determine the actual total payment amount according to the payment results of each period; The total payment amount for each session is determined based on the sum of the estimated total payment amount and the actual total payment amount.
6. The method for estimating the policy renewal rate according to claim 1, characterized in that: After determining the policy renewal rate of the target policy according to the total number of installment payments and the target installment premium, the method further includes: Get the target query task; The query result corresponding to the target query task is determined from the policy renewal rate according to the target query task; wherein the query result is used to represent the renewal rate of the dimension corresponding to the target query task in the policy renewal rate.
7. A device for estimating policy renewal rate, characterized in that: include: The policy data acquisition module is used to acquire the policy data of the target policy; A payment probability output module, used to input the policy data into a payment estimation model to obtain the payment probability output by the payment estimation model; wherein the payment estimation model is trained based on policy sample data, and the policy sample data is pre-acquired sample data; An estimated payment determination module is used to determine an estimated total payment amount according to the payment probability and the policy period premium; wherein the policy period premium refers to the premium corresponding to all periods of the target policy; the estimated total payment amount refers to the sum of payment amounts corresponding to the periods of the target policy that have not reached the payment time; The installment payment determination module is used to determine the installment payment total according to the estimated payment total and the actual payment total; wherein the actual payment total refers to the sum of payment amounts corresponding to the installments of the target insurance policy that have reached the payment time, and the installment payment total refers to the sum of actual payment fees for all installments of the target insurance policy; The renewal rate determination module determines the policy renewal rate of the target policy based on the total number of payments made during each period and the target premium during each period; wherein the target premium during each period refers to the sum of all payable fees for all periods of the target policy.
8. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein: When the processor executes the computer program, it implements the method for estimating the policy renewal rate as described in any one of claims 1 to 6.
9. A non-transitory computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, it implements the method for estimating the policy renewal rate as described in any one of claims 1 to 6.
10. A computer program product, comprising a computer program, characterized in that When the computer program is executed by a processor, it implements the method for estimating the policy renewal rate as described in any one of claims 1 to 6.