Asset level determination method, electronic equipment and computer readable storage medium

Abstract

The invention relates to the technical field of data processing, and particularly provides an asset level determination method, electronic equipment and a computer readable storage medium, the asset level determination method comprises the following steps: obtaining customer label data, the customer label data comprising attribute information, asset information and investment behavior information of a target customer; determining a wallet share of the target customer based on the asset information; and determining the asset level of the target customer based on the attribute information, the asset information, the investment behavior information and the wallet share. In this way, the accuracy of asset level determination can be improved.

Description

Technical Field

[0001] The present application relates to the field of data processing technology, and in particular to a method for determining an asset level, an electronic device, and a computer-readable storage medium. Background Art

[0002] At present, financial institutions need to analyze customer assets and determine the customer's asset level in order to achieve good customer management and service. However, the existing methods for determining asset levels refer to relatively simple customer characteristics. For example, only using the new stock subscription quota to determine the customer's asset level is a one-sided approach that cannot fully and accurately reflect the customer's potential asset size and is prone to omitting important customers, resulting in a decrease in the accuracy of asset level determination, increasing customer asset management risks, and affecting customer management results and subsequent service quality.

[0003] Therefore, how to improve the accuracy of asset grade determination has become a technical problem that needs to be solved urgently. Summary of the invention

[0004] In view of this, the present application provides a method for determining an asset level, an electronic device, and a computer-readable storage medium, which can improve the accuracy of determining a customer's asset level.

[0005] To solve the above technical problems, on the one hand, the present application provides a method for determining an asset level, including: obtaining customer label data, wherein the customer label data includes attribute information, asset information and investment behavior information of the target customer; based on the asset information, determining the wallet share of the target customer; based on the attribute information, asset information, investment behavior information and wallet share, determining the asset level of the target customer.

[0006] According to some embodiments of the present application, the asset information includes the customer's registered assets, the market value of held stocks, and the historical asset peak value, and the wallet share includes the asset allocation rate and the asset utilization rate; the steps of determining the target customer's wallet share based on the asset information include: determining the new stock subscription quota based on the market value of held stocks; determining the asset allocation rate based on the new stock subscription quota and the customer's registered assets; determining the asset utilization rate based on the customer's registered assets, the new stock subscription quota, and the historical asset peak value.

[0007] According to some embodiments of the present application, the steps of determining the asset allocation ratio based on the new stock subscription quota and the customer's registered assets include: calculating the customer's securities assets based on the new stock subscription quota; determining the proportion of the customer's registered assets in the customer's securities assets to generate the asset allocation ratio.

[0008] According to some embodiments of the present application, the steps of determining the asset utilization rate based on the customer's registered assets, new stock subscription quota and historical asset peak value include: calculating the customer's investable assets based on the new stock subscription quota and historical asset peak value; determining the proportion of the customer's registered assets in the customer's investable assets to generate the asset utilization rate.

[0009] According to some embodiments of the present application, the step of determining the asset level of a target customer based on attribute information, asset information, investment behavior information and wallet share includes: inputting the attribute information, asset information, investment behavior information and wallet share into an asset level determination model, and outputting the asset level of the target customer, wherein the asset level determination model is an algorithmic model for determining the asset level based on attribute information, asset information, investment behavior information and wallet share.

[0010] According to some embodiments of the present application, the asset grade determination model is a gradient boosting decision tree and logistic regression model, and the gradient boosting decision tree and logistic regression model include a gradient boosting decision tree processing module and a logistic regression processing module; wherein the gradient boosting decision tree processing module is used to determine the association between attribute information, asset information, investment behavior information and wallet share, convert the attribute information, asset information, investment behavior information and wallet share into high-dimensional features, and input the high-dimensional features into the logistic regression processing module; the logistic regression processing module is used to predict the asset grade based on the high-dimensional features.

[0011] According to some embodiments of the present application, the gradient boosting decision tree processing module includes 30 decision trees, and the maximum depth of each decision tree is 3; the regularization parameter of the logistic regression processing module is set to 0.8, and the elastic network mixing parameter is set to 0.1.

[0012] According to some embodiments of the present application, the gradient boosting decision tree and the logistic regression model are trained using the following steps: obtaining an initial gradient boosting decision tree processing module and an initial logistic regression processing module; obtaining customer label sample data and corresponding wallet share data; training the initial gradient boosting decision tree processing module based on the customer label sample data and the wallet share data to generate a trained gradient boosting decision tree processing module; performing feature conversion on the customer label sample data and the wallet share data using the gradient boosting decision tree processing module to generate a high-dimensional feature set; merging the high-dimensional feature set, the customer label sample data and the wallet share data to obtain training sample data; training the initial logistic regression processing module based on the training sample data to generate a trained logistic regression processing module.

[0013] In a second aspect, an embodiment of the present application provides an electronic device, comprising: a processor; and a memory, in which computer program instructions are stored, wherein when the computer program instructions are executed by the processor, the processor executes the method for determining the asset level as described in the above technical solution.

[0014] In a third aspect, an embodiment of the present application provides a computer-readable storage medium, which stores a computer program. When the computer program is executed by a processor, the processor executes the method for determining the asset level of the above-mentioned technical solution.

[0015] The above technical solution of the present application has at least one of the following beneficial effects: by integrating customer tag data and wallet share, the correlation between different types of characteristics of target customers can be taken into account, so as to comprehensively evaluate the value and development potential of target customers, improve the accuracy of asset level determination, so as to better understand customer asset distribution and investment behavior, optimize customer service and personalized financial product recommendations. Compared with the prior art, the solution proposed in this embodiment can tap 2-4 times more potential high-value customers, greatly expand the high-value customer group, and lay the foundation for good customer management and service of financial institutions.

[0016] In addition, customer label data can be obtained through the existing labeling system to ensure the universality of the above solution within the industry.

[0017] In addition, the algorithm model can realize the automatic determination of customer asset level, reduce manual intervention, improve the accuracy of asset level determination, and improve the efficiency of asset level determination, making business processes more efficient and standardized, able to respond to customer needs more quickly, and optimize customer service experience.

[0018] In addition, by combining the advantages of the GBDT and LR algorithms, the nonlinear modeling capabilities of the GBDT algorithm and the linear modeling capabilities of the LR algorithm are combined to give full play to the high performance advantages of the GBDT-LR model in processing high-dimensional nonlinear data, thereby further improving the accuracy of asset grade determination. In addition, the linear characteristics of the LR algorithm can help users understand which features have the greatest impact on asset grade prediction. The GBDT algorithm has strong robustness to noise data and outliers, and also guarantees the stability of asset grade determination.

[0019] In addition, by optimizing the parameters of the GBDT module and using a large number of decision trees and appropriate tree depth, it is possible to ensure that the GBDT module fully captures the interactions between features and deeply explores the cross-relationships between features. Moreover, by optimizing the parameters of the LR module, setting a larger regularization parameter and a smaller elastic network hybrid parameter, the generalization ability of the model can be maintained, overfitting can be avoided, and the features generated by the GBDT module can be effectively classified, thereby improving the accuracy of asset grade prediction.

[0020] In addition, the initial GBDT module is first trained using customer label sample data and wallet share data to generate a GBDT module that can accurately analyze the association between features. The trained GBDT module is then used to analyze the association between the features in the customer label sample data and wallet share data to generate a high-dimensional feature set. Finally, the initial LR module is trained using the high-dimensional feature set, customer label sample data, and wallet share data to provide richer information dimensions for the training of the LR module, which helps to improve the classification accuracy of the LR module. BRIEF DESCRIPTION OF THE DRAWINGS

[0021] Figure 1 A flowchart of a method for determining an asset level according to an embodiment of the present application;

[0022] Figure 2 A flowchart of the steps of determining the wallet share of a target customer based on asset information in the method for determining the asset level of an embodiment of the present application;

[0023] Figure 3 A schematic diagram of an electronic device according to an embodiment of the present application. DETAILED DESCRIPTION

[0024] The specific implementation methods of the present application will be further described in detail below in conjunction with the accompanying drawings and examples. The following examples are used to illustrate the present application but are not intended to limit the scope of the present application.

[0025] At present, financial institutions (such as securities companies) often need to analyze customer assets and determine the customer's asset level in order to achieve good customer management and service. However, the existing methods for determining asset levels refer to relatively simple customer characteristics. For example, only using the new stock subscription quota to determine the customer's asset level is a one-sided approach that cannot fully and accurately reflect the customer's potential asset size and is prone to omitting important customers, resulting in a decrease in the accuracy of asset level determination, increasing customer asset management risks, and affecting customer management results and subsequent service quality.

[0026] Therefore, how to improve the accuracy of asset grade determination has become a technical problem that needs to be solved urgently.

[0027] In order to solve the above technical problems, this application proposes a method for determining asset levels, which obtains customer tag data and determines the target customer's wallet share based on asset information. Then, the customer tag data and wallet share are used to jointly determine the asset level. This method can take into account the correlation between multiple customer characteristics, more comprehensively evaluate the customer's value and development potential, thereby achieving accurate asset level determination and providing a basis for good customer management and service.

[0028] The following is a detailed description of the method for determining the asset level proposed in this application. The specific steps are as follows: Figure 1 The following steps in the process are shown.

[0029] Step 110: Acquire customer tag data, where the customer tag data includes attribute information, asset information, and investment behavior information of the target customer.

[0030] Specifically, the attribute information may include at least one of age, education level, gender, city, and occupation, which can reflect the personal characteristics and income level of the target customer.

[0031] Asset information may include the client's registered assets, where "registered assets" refers to the assets that the target client has entrusted to a specific financial institution (hereinafter referred to as "this institution") to manage or the assets reported by the client, and "assets" may include but are not limited to investment assets and cash in financial products such as stocks, bonds, and funds. In one example, asset information may also include at least one of the following information: historical asset peaks (such as this year's asset peak, last year's asset peak, the asset peak in the past three years, and the asset peak since account opening), total assets, the market value of stocks held, and the average daily stock assets in the past month.

[0032] Investment behavior information may include at least one of the following information: turnover rate, idle capital rate, stock selection accuracy, investment experience, stock type activity, stock trading frequency, holding period preference, time since the last deposit, number of active days in the past 30 days, average available funds in the past 7 days, number of times financial products have been purchased in the past three months, and investment intention in the Science and Technology Innovation Board, which can reflect the customer's investment preferences.

[0033] In one example, customer label data can be obtained through an existing label system, wherein the label system can generate customer label data using a large number of fact labels and labels that have been statistically calculated based on the fact labels. It can be understood that the label system generating customer label data is a conventional operation in the prior art, and the label system and the process of generating customer label data using the label system are not described in detail and limited here. By obtaining customer label data from an existing label system, the versatility of the method of the present application in the industry can be guaranteed.

[0034] Step 120: Determine the wallet share of the target customer based on the asset information.

[0035] Generally speaking, "wallet share" refers to the proportion of the client's registered assets in the client's total investment amount, which can reflect the target client's trust in the institution and the fund allocation preference. For example, if the target client's total annual investment amount is 1 million yuan, of which 500,000 yuan is invested through the institution, then the institution's wallet share in the target client is 50%.

[0036] Step 130: Determine the asset level of the target customer based on the attribute information, asset information, investment behavior information and wallet share.

[0037] Wallet share can reflect the target customer's trust in the institution and their fund allocation preferences. Attribute information, asset information and investment behavior information can reflect the target customer's income level, financial situation and investment preferences. For example, "30-day active days" can reflect the target customer's trading activity, "average daily stock assets in the past month" reflects the target customer's investment preferences and risk tolerance, and "total assets" reflect the target customer's asset capacity. The higher the total assets, the greater the probability of being classified as a high asset level.

[0038] By combining four different types of features in customer tag data, namely attribute information, asset information, investment behavior information and wallet share, we can have a more comprehensive understanding of the customer's asset status and investment capabilities, thereby determining a more accurate asset level.

[0039] By incorporating customer tag data and wallet share, this embodiment can take into account the correlation between different types of target customer characteristics, thereby comprehensively evaluating the value and development potential of target customers, improving the accuracy of asset level determination, and better understanding customer asset distribution and investment behavior, optimizing customer service and personalized financial product recommendations. Compared with the prior art, the solution proposed in this embodiment can tap 2-4 times more potential high-value customers, greatly expand the high-value customer group, and lay the foundation for good customer management and service for financial institutions. In addition, customer tag data can be obtained through the existing tag system to ensure the universality of the above solution in the industry.

[0040] The above steps 120 and 130 are described in detail below.

[0041] First, in one embodiment, the wallet share may include the asset allocation ratio and the asset utilization ratio, wherein the asset allocation ratio may be used to reflect the proportion of the target customer's assets allocated to the institution, and the asset allocation of the target customer in the institution may be known through the asset allocation ratio. The asset utilization ratio may be used to reflect the proportion of the target customer's assets allocated to the institution to its investable assets, and the asset utilization ratio may be used to know the target customer's asset utilization efficiency in the institution.

[0042] Specifically, if Figure 2 As shown, the above step 120, i.e., the step of determining the wallet share of the target customer based on the asset information, may specifically include:

[0043] Step 121: Determine the new share subscription quota based on the market value of the stocks held.

[0044] The target customer's new stock subscription quota (or new stock subscription quota) can be calculated based on the target customer's stock market value held in all venues in the city. In one example, the new stock subscription quota can be obtained by multiplying the stock market value held by the subscription ratio. Among them, the subscription ratio is affected by the market where the target customer is located and the amount of assets. The specific value can be selected according to the actual situation and is not limited here. It can be understood that other conventional means in the prior art can be used to determine the new stock subscription quota based on the stock market value held, and the calculation method is not limited here.

[0045] Step 122: Determine the asset allocation ratio based on the new share subscription quota and the customer's registered assets.

[0046] In one embodiment, the client's securities assets can be calculated based on the new stock subscription amount, and then the proportion of the client's registered assets in the client's securities assets can be determined to generate an asset allocation ratio.

[0047] Specifically, "customer securities assets" refers to the average daily market value of securities assets held by target customers in the previous 20 trading days. In one example, the new stock subscription quota and the subscription unit (such as 500 shares) can be multiplied to obtain the customer's securities assets. Among them, the subscription unit is affected by the market and asset amount of the target customer. The specific value can be selected according to the actual situation and is not limited here. The total securities assets of the target customer estimated by the new stock subscription quota can measure the proportion of the customer's registered assets to his total securities assets, thereby reflecting the asset allocation of the target customer in this institution.

[0048] It is understandable that other conventional means in the prior art may be used to calculate the customer's securities assets based on the new share subscription amount, and the calculation method is not limited here.

[0049] Step 123: Determine the asset utilization rate based on the customer's registered assets, new stock subscription quota and historical asset peak value.

[0050] In one embodiment, the client's investable assets can be calculated based on the new stock subscription amount and the historical asset peak, and then the proportion of the client's registered assets in the client's investable assets can be determined to generate the asset utilization rate. By calculating the asset utilization rate, it is possible to measure the proportion of the target client's assets allocated in the institution to its investable assets, thereby reflecting the target client's asset utilization efficiency in the institution.

[0051] The above wallet shares (i.e., asset allocation rate and asset utilization rate) reflect the client’s trust in the institution and fund allocation preference, and can serve as an important reference indicator for determining asset grades to improve the accuracy of asset grade determination.

[0052] Specifically, the higher the asset allocation rate and asset utilization rate, the higher the wallet share, which means that the target customer has a higher trust in the institution and the more concentrated the fund allocation preference, which means that the target customer's asset level may be higher and the customer may be more inclined to keep the funds in the institution and invest them.

[0053] The lower the asset allocation rate and asset utilization rate, the lower the wallet share, indicating that the target customer has a lower trust in the institution and the more dispersed the fund allocation is, which may mean that the target customer has a lower asset level and may be more inclined to spread the funds to other institutions or make other investments. However, it also means that the potential value of the customer has not been fully developed, or that the customer may be a potential high-asset customer.

[0054] Therefore, the present application utilizes step 130 to refer to the attribute information, asset information, investment behavior information and wallet share to jointly determine the asset level of the target customer, associate the attribute information, asset information, investment behavior information and wallet share, and conduct a more comprehensive assessment of the target customer, thereby exploring the customer's potential asset scale, improving the accuracy of asset level determination, and discovering potential high-asset-level customers.

[0055] The above step 130, i.e., the step of determining the asset level of the target customer based on the attribute information, asset information, investment behavior information and wallet share, is described in detail below.

[0056] In one embodiment, the attribute information, asset information, investment behavior information and wallet share can be input into the asset level determination model to output the asset level of the target customer, wherein the asset level determination model is an algorithm model for determining the asset level based on the attribute information, asset information, investment behavior information and wallet share. It can be understood that the asset level determination model can be any conventional algorithm model in the technical field that can realize asset level determination, and the type of the algorithm model is not limited here.

[0057] Specifically, the asset level determination model can analyze the complementary relationship between attribute information, asset information, investment behavior information and wallet share. "Complementary relationship" means that the various characteristics complement each other to some extent and jointly provide information about the customer's asset level. Therefore, a comprehensive customer profile can be constructed using attribute information, asset information, investment behavior information and wallet share, thereby accurately predicting the customer's asset level.

[0058] For example, the three characteristics of asset allocation rate, 30-day active days and average daily stock assets in the past month appear simultaneously with high frequency in the prediction process of the asset grade determination model. The three characteristics of asset utilization rate, education level and investment experience appear simultaneously with high frequency in the prediction process of the asset grade determination model, which indicates that there may be a strong "complementary relationship" between them. "Complementary relationship" is also reflected in other different types of feature combinations. Different types of feature combinations can provide richer information than a single feature, thereby improving the accuracy of asset grade prediction. It can be understood that the analysis of feature combinations and complementary relationships between features can be achieved by the asset grade determination model, and this application does not give examples and limitations one by one.

[0059] This embodiment can realize the automatic determination of customer asset level through the algorithm model, reduce manual intervention, improve the accuracy of asset level determination, and improve the efficiency of asset level determination, making business processes more efficient and standardized, and being able to respond to customer needs more quickly and optimize customer service experience.

[0060] In one embodiment, the asset grade determination model may be a gradient boosting decision tree and logistic regression model (Gradient Boosting Decision Tree-Logistic Regression, GBDT-LR), and the GBDT-LR model includes a gradient boosting decision tree processing module (GBDT module) and a logistic regression processing module (LR module). The GBDT module is used to determine the association between attribute information, asset information, investment behavior information and wallet share, convert the attribute information, asset information, investment behavior information and wallet share into high-dimensional features, and input the high-dimensional features into the LR module. The LR module is used to predict the asset grade based on the high-dimensional features.

[0061] Specifically, the GBDT module can determine the association between attribute information, asset information, investment behavior information and wallet share, and convert attribute information, asset information, investment behavior information and wallet share into high-dimensional features through one-hot encoding. The high-dimensional features effectively reveal the association between different features, providing the LR module with richer information dimensions, so that the LR module can predict more accurate asset levels.

[0062] This embodiment can combine the advantages of the GBDT and LR algorithms, combining the nonlinear modeling capabilities of the GBDT algorithm with the linear modeling capabilities of the LR algorithm, and give full play to the high performance advantages of the GBDT-LR model in processing high-dimensional nonlinear data, thereby further improving the accuracy of asset grade determination. In addition, the linear characteristics of the LR algorithm can help users understand which features have the greatest impact on asset grade prediction. The GBDT algorithm has strong robustness to noise data and outliers, and also guarantees the stability of asset grade determination.

[0063] In one embodiment, the GBDT module may include 30 decision trees, and the maximum depth of each decision tree is 3. The regularization parameter (ragParam) of the LR module may be set to 0.8, and the elastic network mixing parameter (ElasticNet mixing parameter) may be set to 0.1.

[0064] This embodiment optimizes the parameters of the GBDT module and uses a large number of decision trees and appropriate tree depths to ensure that the GBDT module fully captures the interactions between features and deeply explores the cross-relationships between features. In addition, by optimizing the parameters of the LR module, setting a larger regularization parameter and a smaller elastic network hybrid parameter, the generalization ability of the model can be maintained, overfitting can be avoided, and the features generated by the GBDT module can be effectively classified to improve the accuracy of asset grade prediction.

[0065] It can be understood that the above 30 decision trees, maximum depth 3, regularization parameter 0.8, and elastic network mixing parameter 0.1 are only examples of model parameters of the GBDT-LR model. In actual application, the above parameters can be adjusted up and down according to needs, and are not limited here.

[0066] In one embodiment, the GBDT-LR model is trained using the following steps: obtaining an initial GBDT module and an initial LR module. Obtaining customer label sample data and corresponding wallet share data. Specifically, the customer label sample data may include attribute sample data, investment behavior sample data, and asset sample data, and the wallet share data is determined based on the asset sample data, wherein the processing steps of "obtaining wallet share data based on asset sample data" are the same as the above steps 121 to 123, and are not repeated here.

[0067] After obtaining the customer label sample data and the corresponding wallet share data, the initial GBDT module can be trained based on the customer label sample data and the wallet share data to generate a trained GBDT module. Then, the GBDT module is used to perform feature conversion on the customer label sample data and the wallet share data to generate a high-dimensional feature set. The high-dimensional feature set, the customer label sample data, and the wallet share data are merged to obtain the training sample data. Finally, the initial LR module is trained based on the training sample data to generate a trained LR module.

[0068] This embodiment first uses customer label sample data and wallet share data to train the initial GBDT module to generate a GBDT module that can accurately analyze the association between features. Then, the trained GBDT module is used to analyze the association between each feature in the customer label sample data and wallet share data to generate a high-dimensional feature set. Finally, the initial LR module is trained using the training sample data obtained by merging the high-dimensional feature set, customer label sample data, and wallet share data, providing a richer information dimension for the training of the LR module, which helps to improve the classification accuracy of the LR module.

[0069] An embodiment of the present application also provides an electronic device, which includes a processor and a memory, wherein the memory stores at least one instruction or at least one program, and the at least one instruction or the at least one program is loaded and executed by the processor to implement the method for determining the asset level provided in the above method embodiment.

[0070] The memory can be used to store software programs and modules. The processor executes various functional applications and data processing by running the software programs and modules stored in the memory. The memory can mainly include a program storage area and a data storage area, wherein the program storage area can store an operating system, application programs required for functions, etc.; the data storage area can store data created according to the use of the device, etc. In addition, the memory can include a high-speed random access memory and can also include a non-volatile memory, such as at least one disk storage device, a flash memory device, or other volatile solid-state storage devices. Accordingly, the memory can also include a memory controller to provide the processor with access to the memory.

[0071] In a specific embodiment, Figure 3 A schematic diagram of the structure of an electronic device for implementing the embodiment of the present application is shown, and the electronic device may be a computer terminal, a mobile terminal or other devices.

[0072] like Figure 3As shown, an embodiment of the present application provides an electronic device 300, including: a processor 301 and a memory 302, wherein computer program instructions are stored in the memory 302, wherein when the computer program instructions are executed by the processor, the processor 301 executes the method for determining the asset level as described in the above embodiment.

[0073] Furthermore, if Figure 3 As shown, the electronic device 300 further includes a network interface 303 , an input device 304 , a hard disk 305 , and a display device 306 .

[0074] The above-mentioned interfaces and devices can be interconnected through a bus architecture. The bus architecture can be a bus and bridge that can include any number of interconnected buses. Specifically, one or more central processing units (CPUs) represented by processor 301 and various circuits of one or more memories represented by memory 302 are connected together. The bus architecture can also connect various other circuits such as peripheral devices, voltage regulators, and power management circuits together. It can be understood that the bus architecture is used to achieve connection and communication between these components. In addition to the data bus, the bus architecture also includes a power bus, a control bus, and a status signal bus, which are all well known in the art, so they are not described in detail herein.

[0075] The network interface 303 can be connected to a network (such as the Internet, a local area network, etc.), obtain relevant data from the network, and save it in the hard disk 305.

[0076] The input device 304 can receive various instructions input by the operator and send them to the processor 301 for execution. The input device 304 can include a keyboard or a pointing device, such as a mouse, a trackball, a touch pad or a touch screen.

[0077] The display device 306 can display the result obtained by the processor 301 executing the instruction.

[0078] The memory 302 is used to store programs and data necessary for the operation of the operating system, as well as data such as intermediate results during the calculation process of the processor 301.

[0079] It is understood that the memory 302 in the embodiment of the present application can be a volatile memory or a non-volatile memory, or can include both volatile and non-volatile memory. Among them, the non-volatile memory can be a read-only memory (ROM), a programmable read-only memory (PROM), an erasable programmable read-only memory (EPROM), an electrically erasable programmable read-only memory (EEPROM) or a flash memory. The volatile memory can be a random access memory (RAM), which is used as an external cache. The memory 302 of the apparatus and method described herein is intended to include, but is not limited to, these and any other suitable types of memory.

[0080] In some implementations, the memory 302 stores the following elements, executable modules or data structures, or a subset thereof, or an extended set thereof: an operating system 3021 and an application program 3022 .

[0081] The operating system 3021 includes various system programs, such as a framework layer, a core library layer, a driver layer, etc., which are used to implement various basic services and process hardware-based tasks. The application 3022 includes various application programs, such as a browser, etc., which are used to implement various application services. The program for implementing the method of the embodiment of the present application can be included in the application 3022.

[0082] The processor 301 , when calling and executing the application and data stored in the memory 302 , specifically, the program or instruction stored in the application 3022 , executes the method for determining the asset level as described in the above embodiment.

[0083] The method disclosed in the above embodiment of the present application can be applied to the processor 301, or implemented by the processor 301. The processor 301 may be an integrated circuit chip with signal processing capabilities. In the implementation process, each step of the above method can be completed by the hardware integrated logic circuit or software instructions in the processor 301. The above-mentioned processor 301 can be a general-purpose processor, a digital signal processor (DSP), an application-specific integrated circuit (ASIC), a readily available programmable gate array (FPGA) or other programmable logic devices, discrete gates or transistor logic devices, discrete hardware components, and can implement or execute the methods, steps and logic block diagrams disclosed in the embodiments of the present application. The general-purpose processor can be a microprocessor or the processor can also be any conventional processor, etc. The steps of the method disclosed in the embodiment of the present application can be directly embodied as a hardware decoding processor to execute, or the hardware and software modules in the decoding processor are combined and executed. The software module can be located in a mature storage medium in the field such as a random access memory, a flash memory, a read-only memory, a programmable read-only memory or an electrically erasable programmable memory, a register, etc. The storage medium is located in the memory 302, and the processor 301 reads the information in the memory 302 and completes the steps of the above method in combination with its hardware.

[0084] It is understood that the embodiments described herein can be implemented in hardware, software, firmware, middleware, microcode or a combination thereof. For hardware implementation, the processing unit can be implemented in one or more application specific integrated circuits (ASICs), digital signal processors (DSPs), digital signal processing devices (DSPDs), programmable logic devices (PLDs), field programmable gate arrays (FPGAs), general purpose processors, controllers, microcontrollers, microprocessors, other electronic units for performing the functions described in the present application or a combination thereof.

[0085] For software implementation, the techniques described herein can be implemented by modules (e.g., procedures, functions, etc.) that perform the functions described herein. The software code can be stored in a memory and executed by a processor. The memory can be implemented in the processor or outside the processor.

[0086] In addition, an embodiment of the present application further provides a computer-readable storage medium, wherein the computer-readable storage medium stores a computer program. When the computer program is executed by a processor, the processor executes the method for determining the asset level as described in the above embodiment.

[0087] In the several embodiments provided in the present application, it should be understood that the disclosed methods and devices can be implemented in other ways. For example, the device embodiments described above are only schematic. For example, the division of the units is only a logical function division. There may be other division methods in actual implementation, such as multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, indirect coupling or communication connection of devices or units, which can be electrical, mechanical or other forms.

[0088] In addition, each functional unit in each embodiment of the present application may be integrated into one processing unit, or each unit may be physically included separately, or two or more units may be integrated into one unit. The above-mentioned integrated unit may be implemented in the form of hardware or in the form of hardware plus software functional units.

[0089] The above-mentioned integrated unit implemented in the form of a software functional unit can be stored in a computer-readable storage medium. The above-mentioned software functional unit 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 some steps of the sending and receiving method described in each embodiment of the present application. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (Read-Only Memory, referred to as ROM), random access memory (RandomAccess Memory, referred to as RAM), disk or optical disk and other media that can store program codes.

[0090] The above is a preferred embodiment of the present application. It should be pointed out that for ordinary technicians in this technical field, several improvements and modifications can be made without departing from the principles described in the present application. These improvements and modifications should also be regarded as the scope of protection of the present application.

Claims

1. A method for determining an asset grade, characterized in that: The steps include: Acquire customer tag data, wherein the customer tag data includes attribute information, asset information, and investment behavior information of the target customer; Based on the asset information, determining the wallet share of the target customer; The asset level of the target customer is determined based on the attribute information, the asset information, the investment behavior information and the wallet share.

2. The method for determining asset level according to claim 1, characterized in that: The asset information includes the customer's registered assets, the market value of the stocks held, and the historical asset peak value; the wallet share includes the asset allocation rate and the asset utilization rate; the step of determining the target customer's wallet share based on the asset information includes: Determine the new stock subscription quota based on the market value of the stocks held; Determining the asset allocation ratio based on the new share subscription amount and the customer's registered assets; The asset utilization rate is determined based on the customer's registered assets, the new stock subscription amount and the historical asset peak value.

3. The method for determining asset level according to claim 2, characterized in that: The step of determining the asset allocation ratio based on the new share subscription amount and the customer's registered assets includes: Based on the new share subscription amount, the client's securities assets are calculated; Determine the proportion of the client's registered assets in the client's securities assets and generate the asset allocation ratio.

4. The method for determining asset level according to claim 2, characterized in that: The step of determining the asset utilization rate based on the customer's registered assets, the new share subscription amount and the historical asset peak value includes: Based on the new share subscription amount and the historical asset peak, the client's investable assets are calculated; Determine the proportion of the client's registered assets in the client's investable assets and generate the asset utilization rate.

5. The method for determining asset level according to claim 1, characterized in that: The step of determining the asset level of the target customer based on the attribute information, the asset information, the investment behavior information and the wallet share comprises: The attribute information, the asset information, the investment behavior information and the wallet share are input into an asset grade determination model, and the asset grade of the target customer is output, wherein the asset grade determination model is an algorithmic model for determining the asset grade based on the attribute information, the asset information, the investment behavior information and the wallet share.

6. The method for determining asset level according to claim 5, characterized in that: The asset grade determination model is a gradient boosting decision tree and logistic regression model, and the gradient boosting decision tree and logistic regression model includes a gradient boosting decision tree processing module and a logistic regression processing module; The gradient boosting decision tree processing module is used to determine the association between the attribute information, the asset information, the investment behavior information and the wallet share, convert the attribute information, the asset information, the investment behavior information and the wallet share into high-dimensional features, and input the high-dimensional features into the logistic regression processing module; the logistic regression processing module is used to predict the asset level based on the high-dimensional features.

7. The method for determining asset level according to claim 6, characterized in that: The gradient boosting decision tree processing module includes 30 decision trees, and the maximum depth of each decision tree is 3; the regularization parameter of the logistic regression processing module is set to 0.8, and the elastic network mixing parameter is set to 0.

1.

8. The method for determining asset level according to claim 6, characterized in that: The gradient boosting decision tree and logistic regression model are trained using the following steps: Obtain an initial gradient boosting decision tree processing module and an initial logistic regression processing module; Obtain customer tag sample data and corresponding wallet share data; Based on the customer label sample data and the wallet share data, the initial gradient boosting decision tree processing module is trained to generate a trained gradient boosting decision tree processing module; Using the gradient boosting decision tree processing module to perform feature conversion on the customer label sample data and the wallet share data to generate a high-dimensional feature set; Merging the high-dimensional feature set, the customer label sample data and the wallet share data to obtain training sample data; The initial logistic regression processing module is trained based on the training sample data to generate a trained logistic regression processing module.

9. An electronic device, characterized in that: include: processor; and a memory having computer program instructions stored therein, When the computer program instructions are executed by the processor, the processor executes the method for determining the asset grade according to any one of claims 1 to 8.

10. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores computer program instructions, and when the computer program instructions are executed by a processor, the processor executes the method for determining an asset grade according to any one of claims 1 to 8.

Images