Financial data processing method, device, medium, electronic equipment and program product

Abstract

This application discloses a method, apparatus, medium, electronic device, and program product for processing financial data. Relating to the field of artificial intelligence, the method includes: acquiring a first dataset of a first object, wherein the first dataset includes: attribute data of the first object and M types of financial data generated by the first object's purchase of financial products during a historical period; based on the first dataset, selecting a first prediction model from N mixture models, wherein the model type of the N mixture models includes: a stochastic slope intercept model; inputting the first dataset into the first prediction model, and outputting first financial data, wherein the first financial data includes: financial data related to the first object's target financial account in a financial institution, and the financial data includes: consumption data. This application solves the problem of low accuracy in related technologies that use a single regression equation to output consumption data of all customers' financial accounts.

Description

Technical Field

[0001] This application relates to the field of artificial intelligence, and more specifically, to a method, apparatus, medium, electronic device, and program product for processing financial data. Background Technology

[0002] In related technologies, the classic multiple linear regression model, or stepwise regression model, uses fixed independent variables as input to output the total spending amount of a financial account (e.g., credit card). The parameters estimated using the same batch of data are also fixed values. However, different individual customers have different spending habits. Some customers are influenced by different factors and at different times. Using the same regression equation can only accurately output the total credit card spending amount of the customer as a whole, but it cannot distinguish the differences between individual customers. Therefore, the accuracy of outputting the total spending amount of a single customer's financial account using a single regression equation is low.

[0003] There is currently no effective solution to the problem of low accuracy in using a single regression equation to output consumption data for all customers' financial accounts in related technologies. Summary of the Invention

[0004] The main purpose of this application is to provide a method, apparatus, medium, electronic device and program product for processing financial data, so as to solve the problem of low accuracy in related technologies that use a single regression equation to output consumption data of all customers' financial accounts.

[0005] To achieve the above objectives, according to one aspect of this application, a method for processing financial data is provided. The method includes: acquiring a first dataset of a first object, wherein the first dataset includes: attribute data of the first object and M types of financial data generated by the first object purchasing financial products during a historical time period, where M is a positive integer; based on the first dataset, selecting a first prediction model from N hybrid models, wherein the model types of the N hybrid models include: a stochastic slope intercept model, wherein the stochastic slope intercept model includes: a model constructed based on a stochastic slope and a stochastic intercept, where N is a positive integer; inputting the first dataset into the first prediction model, and outputting first financial data, wherein the first financial data includes: financial data related to the first object's target financial account in a financial institution, wherein the financial data includes: consumption data.

[0006] Further, based on the first dataset, selecting a first prediction model from N hybrid models includes: obtaining N second datasets, wherein the N second datasets include: attribute data of each of the N second objects and M types of financial data generated by the purchase of financial products by the second object; matching the first dataset and the N second datasets to obtain a first matching result, wherein the first matching result is used to indicate the second dataset among the N second datasets that successfully matches the first dataset; and determining the hybrid model associated with the second dataset indicated by the matching result as the first prediction model.

[0007] Further, the N hybrid models are obtained by: determining S linear models, wherein the model types of the S linear models include at least one of the following: a stochastic slope model, a stochastic intercept model, and a stochastic slope-intercept model, where S is a positive integer; performing data fitting on the N second datasets using each of the linear models to obtain the fitting result of each linear model, and determining the N hybrid models based on the fitting result, wherein the data fitting is used to determine the parameter values ​​of the unknown parameters in the linear model.

[0008] Further, based on the fitting results, N hybrid models are determined, including: verifying each fitting result based on a target verification strategy to obtain a verification result for each linear model, wherein the target verification strategy includes at least one of the following: significance test, regression coefficient test, and goodness-of-fit test; if the verification results of S linear models all meet the preset conditions, N hybrid models are determined based on the fitting results of the random slope intercept model.

[0009] Furthermore, after determining N hybrid models based on the fitting results of the random slope intercept model, the method further includes: extracting the random slope and random intercept of each of the N hybrid models to obtain a third dataset; and constructing a second prediction model based on the third dataset, wherein the second prediction model includes a model with random slope and random intercept as dependent variables and financial data generated by the target financial account as independent variables.

[0010] Furthermore, the financial data of the target financial account includes: the total transaction amount of the target financial account during the target time period. After constructing the second prediction model based on the third dataset, the method further includes: obtaining the financial data of the target financial account of the third object in the financial institution to obtain the second financial data; inputting the second financial data into the second prediction model and outputting a target slope and a target intercept, wherein the target slope is used to determine the degree of influence of the second financial data on the attribute data of the third object or the purchase of financial products by the third object, and the target intercept is used to determine the lowest total transaction amount of the third object at multiple time points during the target time period.

[0011] Furthermore, after inputting the first dataset into the first prediction model and outputting the first financial data, the method further includes: obtaining a preset mapping rule, matching the first financial data with the preset mapping rule to obtain a second matching result, wherein the preset mapping rule includes: a mapping relationship between the financial data of the target financial account and T types of financial products, where T is a positive integer; and determining the target financial product based on the second matching result, wherein the target financial product includes: financial products to be recommended to the first object.

[0012] To achieve the above objectives, according to another aspect of this application, a financial data processing apparatus is provided. The apparatus includes: a first acquisition unit, configured to acquire a first dataset of a first object, wherein the first dataset includes: attribute data of the first object and M types of financial data generated by the first object purchasing financial products during a historical time period, where M is a positive integer; a filtering unit, configured to filter a first prediction model from N hybrid models based on the first dataset, wherein the model types of the N hybrid models include: a stochastic slope intercept model, wherein the stochastic slope intercept model includes: a model constructed based on a stochastic slope and a stochastic intercept, where N is a positive integer; and a first processing unit, configured to input the first dataset into the first prediction model and output first financial data, wherein the first financial data includes: financial data related to the first object's target financial account in a financial institution, wherein the financial data includes: consumption data.

[0013] Further, the filtering unit includes: an acquisition subunit for acquiring N second datasets, wherein the N second datasets include: attribute data of each of the N second objects and M types of financial data generated by the purchase of financial products by the second object; a matching subunit for matching the first dataset and the N second datasets to obtain a first matching result, wherein the first matching result is used to indicate the second dataset among the N second datasets that successfully matches the first dataset; and a first determination subunit for determining the hybrid model associated with the second dataset indicated by the matching result as the first prediction model.

[0014] Further, the N hybrid models are obtained through the following sub-units: a second determining sub-unit, used to determine S linear models, wherein the model types of the S linear models include at least one of the following: a stochastic slope model, a stochastic intercept model, and a stochastic slope-intercept model, where S is a positive integer; and a fitting sub-unit, used to perform data fitting on the N second datasets using each of the linear models to obtain the fitting result of each of the linear models, and based on the fitting result, to determine the N hybrid models, wherein the data fitting is used to determine the parameter values ​​of the unknown parameters in the linear model.

[0015] Further, the fitting subunit includes: a verification module, used to verify each fitting result based on a target verification strategy to obtain a verification result for each linear model, wherein the target verification strategy includes at least one of the following: significance test, regression coefficient test and goodness-of-fit test; and a determination module, used to determine N mixed models based on the fitting result of the random slope intercept model when the verification results of the S linear models all meet the preset conditions.

[0016] Furthermore, the financial data processing device further includes: an extraction unit, used to extract the random slope and random intercept of each of the N mixed models after determining N mixed models based on the fitting results of the random slope intercept model, to obtain a third dataset; and a construction subunit, used to construct a second prediction model based on the third dataset, wherein the second prediction model includes: a model with random slope and random intercept as dependent variables and financial data generated by the target financial account as independent variables.

[0017] Further, the financial data of the target financial account includes: the total transaction amount of the target financial account during the target time period. The financial data processing device further includes: a second acquisition unit, used to acquire the financial data of the target financial account of the third object in the financial institution after constructing a second prediction model based on the third dataset, to obtain the second financial data; and a second processing unit, used to input the second financial data into the second prediction model and output a target slope and a target intercept, wherein the target slope is used to determine the degree of influence of the second financial data on the attribute data of the third object or the purchase of financial products by the third object, and the target intercept is used to determine the lowest total transaction amount of the third object at multiple time points during the target time period.

[0018] Furthermore, the financial data processing device further includes: a third acquisition unit, configured to, after inputting the first dataset into the first prediction model and outputting the first financial data, acquire a preset mapping rule, match the first financial data with the preset mapping rule, and obtain a second matching result, wherein the preset mapping rule includes: a mapping relationship between the financial data of the target financial account and T types of financial products, where T is a positive integer; and a determination unit, configured to, based on the second matching result, determine a target financial product, wherein the target financial product includes: a financial product to be recommended to the first object.

[0019] According to another aspect of this application, a computer-readable storage medium is provided, the computer-readable storage medium including a stored executable program, wherein, when the executable program is executed, it controls the device where the computer-readable storage medium is located to perform the method for processing the financial data.

[0020] According to another aspect of this application, an electronic device is provided, comprising: a memory storing an executable program; and a processor for running the program, wherein the program, when running, performs the method for processing the financial data.

[0021] According to another aspect of this application, a computer program product is provided, including computer instructions that, when executed by a processor, implement the steps of the financial data processing method.

[0022] In this application, a first dataset of a first object is obtained, comprising: attribute data of the first object and M types of financial data generated by the first object's purchase of financial products during a historical period, where M is a positive integer. Based on the first dataset, a first prediction model is selected from N mixture models, wherein the model types of the N mixture models include: a stochastic slope intercept model, and the stochastic slope intercept model includes: a model constructed based on a stochastic slope and a stochastic intercept, where N is a positive integer. The first dataset is input into the first prediction model, and the first financial data is output, comprising: financial data related to the first object's target financial account in a financial institution, including: consumption data. This solves the technical problem of low accuracy in related technologies that use a single regression equation to output consumption data of all customers' financial accounts. In this invention, based on the dataset of the first object, a first prediction model associated with the first object is selected from N mixture models, and the first prediction model is used to output the consumption data of the first object's target financial account. This avoids the situation in related technologies where a single regression equation is used to output consumption data of all customers' target financial accounts, which cannot distinguish individual differences, thereby achieving the technical effect of accurately determining the consumption data of a single customer's financial account. Attached Figure Description

[0023] The accompanying drawings, which form part of this application, are used to provide a further understanding of this application. The illustrative embodiments and descriptions of this application are used to explain this application and do not constitute an undue limitation of this application. In the drawings:

[0024] Figure 1 A hardware structure block diagram of a computer terminal for implementing a method for processing financial data is shown.

[0025] Figure 2 This is a flowchart of a financial data processing method provided according to an embodiment of this application;

[0026] Figure 3 This is an individual and group relationship diagram provided according to the embodiments of this application;

[0027] Figure 4 This is a schematic diagram of the first coordinate system provided according to an embodiment of this application;

[0028] Figure 5 This is a schematic diagram of the second coordinate system provided according to an embodiment of this application;

[0029] Figure 6 This is a schematic diagram of a financial data processing apparatus provided according to an embodiment of this application;

[0030] Figure 7 This is a structural block diagram of an electronic device according to an embodiment of this application. Detailed Implementation

[0031] To enable those skilled in the art to better understand the present application, the technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present application, and not all embodiments. Based on the embodiments in the present application, all other embodiments obtained by those of ordinary skill in the art without creative effort should fall within the scope of protection of the present application.

[0032] It should be noted that the terms "first," "second," etc., in the specification, claims, and accompanying drawings of this application are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments of this application described herein can be implemented in orders other than those illustrated or described herein. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover non-exclusive inclusion; for example, a process, method, system, product, or apparatus that comprises a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or apparatus.

[0033] It should be noted that the information collected in this application (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for display, data used for analysis, etc.) are information and data authorized by the user or fully authorized by all parties. Furthermore, the collection, storage, use, processing, transmission, provision, disclosure, and application of this data all comply with relevant laws, regulations, and standards, necessary confidentiality measures have been taken, and they do not violate public order and good morals. Corresponding access points are provided for users to choose to authorize or refuse. For example, interfaces are set up between this system and relevant users or organizations, providing users with corresponding access points to choose to agree to or refuse automated decision-making results; if the user chooses to refuse, the process proceeds to the expert decision-making stage.

[0034] Example 1

[0035] According to an embodiment of this application, a method embodiment for processing financial data is also provided. It should be noted that the steps shown in the flowchart in the accompanying drawings can be executed in a computer system such as a set of computer-executable instructions. Furthermore, although a logical order is shown in the flowchart, in some cases, the steps shown or described may be executed in a different order than that shown here.

[0036] The method embodiment provided in Embodiment 1 of this application can be executed on a mobile terminal, computer terminal, or similar computing device. Figure 1A hardware block diagram of a computer terminal (or mobile device) for implementing a method for processing financial data is shown. Figure 1 As shown, the computer terminal 10 (or mobile device) may include one or more processors 102 (shown as 102a, 102b, ..., 102n in the figure) 102 (processor 102 may include, but is not limited to, a microprocessor MCU or a programmable logic device FPGA, etc.), a memory 104 for storing data, and a transmission device 106 for communication functions. In addition, it may also include: a display, an input / output interface (I / O interface), a universal serial bus (USB) port (which may be included as one of the ports of a BUS bus), a network interface, a power supply, and / or a camera. Those skilled in the art will understand that... Figure 1 The structure shown is for illustrative purposes only and does not limit the structure of the aforementioned electronic device. For example, computer terminal 10 may also include... Figure 1 The more or fewer components shown, or having the same Figure 1 The different configurations shown.

[0037] It should be noted that the aforementioned one or more processors 102 and / or other data processing circuits are generally referred to herein as "data processing circuits". These data processing circuits may be embodied, in whole or in part, in software, hardware, firmware, or any other combination thereof. Furthermore, the data processing circuits may be a single, independent processing module, or may be integrated, in whole or in part, into any other element within the computer terminal 10 (or mobile device). As involved in the embodiments of this application, the data processing circuits serve as a processor control mechanism (e.g., selection of a variable resistor termination path connected to an interface).

[0038] The memory 104 can be used to store software programs and modules of application software, such as the program instructions / data storage device corresponding to the financial data processing method in this embodiment. The processor 102 executes various functional applications and data processing by running the software programs and modules stored in the memory 104, thereby realizing the aforementioned financial data processing method. The memory 104 may include high-speed random access memory, and may also include non-volatile memory, such as one or more magnetic storage devices, flash memory, or other non-volatile solid-state memory. In some instances, the memory 104 may further include memory remotely located relative to the processor 102, and these remote memories can be connected to the computer terminal 10 via a network. Examples of such networks include, but are not limited to, the Internet, corporate intranets, local area networks, mobile communication networks, and combinations thereof.

[0039] The transmission device 106 is used to receive or send data via a network. Specific examples of the network described above may include a wireless network provided by the communication provider of the computer terminal 10. In one example, the transmission device 106 includes a Network Interface Controller (NIC), which can connect to other network devices via a base station to communicate with the Internet. In another example, the transmission device 106 may be a Radio Frequency (RF) module, used for wireless communication with the Internet.

[0040] The display can be, for example, a touchscreen liquid crystal display (LCD) that allows the user to interact with the user interface of the computer terminal 10 (or mobile device).

[0041] Under the aforementioned operating environment, this application provides the following: Figure 2 The method for processing financial data is shown. Figure 2 This is a flowchart of a financial data processing method according to Embodiment 1 of this application.

[0042] Step S201: Obtain the first dataset of the first object, wherein the first dataset includes: attribute data of the first object and M types of financial data generated by the first object in purchasing financial products during historical time periods, where M is a positive integer.

[0043] The attribute data of the first object mentioned above may include, but is not limited to, basic information such as the first object's occupation, employer, department, work location (e.g., city), salary, and years of service. The first object may include objects that have already established target financial accounts with financial institutions. The first dataset may include, in part, the attribute data of the first object and M types of financial data generated from the first object's purchases of financial products over a historical period. The M types of financial data may include, for example, consumption data generated from payments made using the first object's target financial account for the purchase of financial products.

[0044] Step S202: Based on the first dataset, select the first prediction model from N mixture models. The model types of the N mixture models include: random slope intercept model, which includes: a model built based on random slope and random intercept, where N is a positive integer.

[0045] The aforementioned N hybrid models are used to analyze financial data related to target financial accounts of different objects, such as total consumption over a historical period.

[0046] In this embodiment, a second dataset of objects involved in each hybrid model can be obtained. The second dataset includes: attribute data of the second object and M types of financial data generated by the second object purchasing financial products. The model type of the hybrid model mentioned above can include: a stochastic slope intercept model. The hybrid model can include multiple variables, a stochastic slope, and a stochastic intercept.

[0047] By matching the first dataset and the second dataset, the first prediction model can be selected from N mixed models based on the matching results.

[0048] Step S203: Input the first dataset into the first prediction model and output the first financial data, wherein the first financial data includes: financial data related to the target financial account of the first object in the financial institution, and the financial data includes: consumption data.

[0049] The aforementioned first financial data may include: financial data related to the target financial account of the first object in a financial institution. The financial data includes: consumption data (e.g., total consumption amount). The aforementioned target financial account may include a credit card account. In this embodiment, the first dataset can be input into the first prediction model, and the variables in the first prediction model can be replaced with the data in the first dataset to calculate the first financial data. It should be noted that, in order to avoid the inability to calculate non-numerical data, in this embodiment, the non-numerical data can also be converted into numerical data. The conversion method may include: conversion using one-hot encoding.

[0050] In this embodiment, through the above steps, based on the dataset of the first object, a first prediction model associated with the first object is selected from N hybrid models. The consumption data of the target financial account of the first object is output using the first prediction model. This avoids the situation in related technologies where a single regression equation is used to output the consumption data of the target financial accounts of all customers, which cannot distinguish individual differences. This achieves the technical effect of accurately determining the consumption data of the financial account of a single customer, and solves the technical problem of low accuracy in related technologies where a single regression equation is used to output the consumption data of the financial accounts of all customers.

[0051] Optionally, in the financial data processing method provided in this application embodiment, the first prediction model is selected from N hybrid models based on the first dataset, including: obtaining N second datasets, wherein the N second datasets include: attribute data of each of the N second objects and M types of financial data generated by the purchase of financial products by the second object; matching the first dataset and the N second datasets to obtain a first matching result, wherein the first matching result is used to indicate the second dataset among the N second datasets that successfully matches the first dataset; and determining the hybrid model associated with the second dataset indicated by the matching result as the first prediction model.

[0052] The above N mixed models can include independent variables with fixed effects and independent variables with random effects. The parameters (i.e., fixed parameters) of the independent variables with fixed effects in the N mixed models are the same, while the parameters (random slope and random intercept) of the independent variables with random effects in the N mixed models can be different.

[0053] In this embodiment, a first dataset and N second datasets can be matched. For example, the content corresponding to random variables (i.e., random effects) in the first dataset can be matched with the content corresponding to random variables in the N second datasets. For example, if the first dataset indicates that the working city corresponding to the first object is A and the occupation is B, and a certain second dataset indicates that the working city corresponding to the second object is also A and the occupation is also B, then it can be determined that the first dataset and the second dataset are successfully matched. Then, the mixture model associated with the second dataset can be determined as the first prediction model.

[0054] In this embodiment, different objects can input corresponding mixture models according to different random effects, and output the total consumption amount of the target financial account of the object, thereby achieving the technical effect of improving the calculation accuracy of the total consumption amount of the target financial account.

[0055] Optionally, in the financial data processing method provided in this application embodiment, N hybrid models are obtained in the following manner: determining S linear models, wherein the model types of the S linear models include at least one of the following: stochastic slope model, stochastic intercept model, and stochastic slope-intercept model, where S is a positive integer; using each linear model to fit data to N second datasets to obtain the fitting result of each linear model, and based on the fitting result, determining N hybrid models, wherein the data fitting is used to determine the parameter values ​​of the unknown parameters in the linear model.

[0056] In this embodiment, the model types of the S linear models include, but are not limited to: general linear model, stochastic slope model, stochastic intercept model, and stochastic slope-intercept model.

[0057] General linear model: y i=β0+β1x i +ε i

[0058] Random intercept model: y i =β0+γ0+β1x i +ε i

[0059] Stochastic slope model: y i =β0+(β1+γ1)x i +ε i

[0060] Stochastic intercept and stochastic slope model: y i =β0+γ0+(β1+γ1)x i +ε i

[0061] Where, β i and γ i ε represents the estimated parameter. i This represents random error.

[0062] In this embodiment, attribute information of multiple customers (corresponding to N second objects) and consumption history information of the target financial account (M types of financial data generated by the second objects purchasing financial products) can be collected. The customer information data can be processed and model fitting can be used to output the total credit card consumption Y of each customer based on the model.

[0063] Input data: N customers (e.g., 61 customers) with independent variables X1, X2, X3, X4, X5, X6. Each customer has ni (3 or 4) observations: total credit card spending Y and independent variables X2, X3. These ni different observations are recorded at different times t (1 to 4 weeks). The number of observations, observation intervals, and the number of measurements may vary for each customer. This differs from classic time series models, which require a large number of observations with equal intervals. Output: Predict the total credit card spending Y for each customer, calculating fixed parameters b0, b1, b2, b3, b4, b5, b6, and a random parameter, random intercept γ. N0 and random slope γ N2 and γ N3 The following explains the process of fitting the dataset using a linear model:

[0064] 1. General linear model:

[0065] A general linear model is used: y = Xβ + ε, i = 1, 2, ..., N. The dataset is fitted using the maximum likelihood method to obtain the estimated parameter β and the model variance. Statistical validation is then performed, including significance tests for the regression equation, regression coefficient tests, and goodness-of-fit tests. A multiple linear model is also used: N customers with independent variables X1, X2, X3, X4, X5, X6, and the total credit card spending Y of each customer.

[0066]

[0067] It can also be expressed as:

[0068]

[0069] If the linear model is a stochastic intercept-slope model: y i =X i β i +Z i γ i +ε i Let i = 1, 2, ..., N. By fitting N second datasets, we can obtain the above N mixture models, where the estimated parameter β... i For fixed parameters, estimate the parameter γ. i For random parameters (i.e., random slope and random intercept), when the estimated parameters of N mixture models have not been calculated, it can be represented as:

[0070]

[0071] The dependent variables X1, X2, X3, X4, X5, and X6 of N customers are fixed effects. The independent variables X2 and X3 are added to the model as random effects. Each customer has ni observations: total credit card spending Y and independent variables X2 and X3. These ni different observations are recorded at different times t. i0 .

[0072] For each client, a separate random intercept and random slope model is built, resulting in different intercept estimates (random intercept) and slope estimates (random slope). The intercept and slope are estimated individually for each client.

[0073] If the coefficients of the fixed effects estimate are (x1 values ​​are all 1, the system cannot be estimated, default β1 = 0):

[0074] β0=1.0960, β2=-2.1106, β3=0.4291, β4=0.1790, β5=-0.5643, β6=1.2720

[0075] If the coefficients (or parameters) of the random effects estimate of the mixed model (y1) for the first second object are:

[0076] γ 20 =-5.1817,γ 21 =-1.8940,γ 23 =1.2389

[0077] The hybrid model of the first second object is:

[0078] y1=(β 10 +β 11 x 11 +...+β 16 x 16 )+(γ 10 +γ 12 x 12 +γ 13 x 13 )+ε1

[0079] y1=(1.0960 -2.1106x 12 +0.4291x 13 +0.1790x 14 -0.5643x 15 +1.2720x 16 )

[0080] +(-5.1817-1.8940x 12 +1.2389x 13 )

[0081] If the hybrid model of the second object corresponds to y2, γ 20 =-4.3688,γ 21 =-0.3656,γ 23 =0.0622, then the hybrid model of the second object is:

[0082] y2=(β 20 +β 21 x 21 +...+β 26 x 26 )+(γ 20 +γ 22 x 22 +γ 23 x 23 )+ε2

[0083] y2=(1.0960-2.1106x 22 +0.4291x 23 +0.1790x 24 -0.5643x 25 +1.2720x 26 )+(-4.3688-0.3656x22 +0.0622x 23 )

[0084] It's important to note that random intercepts and slopes allow for differences in rank (intercept) and association (slope) between individuals. For a given individual (unit), even if repeated measures data may exist, there is only one intercept and one slope. Therefore, although random intercepts and slopes are derived from repeated test data (individual stratum), the intercepts and slopes themselves are still repeated measures data at the population stratum.

[0085] In this embodiment, when a random effects mixture model is used for fitting, it includes:

[0086] (1) Observed values ​​y of random effects i , ε i They are all independent random variables. ε i It is an unobservable random error.

[0087] (2) Observed value y i and fixed effects independent variable X i Random effects independent variable Z i It exhibits a clear linear relationship.

[0088] (3) Random errors ε1, ε2, ..., ε n For samples that are independent and identically distributed, ε has a linear relationship with ε. i ~N(0,σ 2 ), i = 1, 2, ..., n.

[0089] (4) Random effect independent variable Z i It follows a normal distribution.

[0090] (5) Observed value y i It follows the expected distribution, normal distribution, Poisson distribution, etc.

[0091] The general form of a linear random effects mixture model (corresponding to a mixture model):

[0092] y i =X i β i +Z i γ i +ε i i = 1, 2, ..., N

[0093] Among them, y i For n i ×r-dimensional, X i For n i ×p-dimensional, β i For p×r dimensions, Z i For ni ×q-dimensional, γ i For q×r dimensions, ε i For n i ×r dimension.

[0094]

[0095] For all i, ε i ~N(0,Σ), and independent of γ i , usually X i and Z i The first column is a constant, Z i It contains X i A subset of , which is to be estimated as β i ,Σ,ψ。 X in the formula i β i For the fixed utility component, and Z i γ i This represents the random utility component. If the data is written in the following form...

[0096]

[0097] as well as:

[0098]

[0099] Model y i =X i β i +Z i γ i +ε i (i = 1, 2, ..., N) can be written as:

[0100] Y = Xβ + Zγ + ε.

[0101] Where Y is n i In the ×r-dimensional case, r = 1, for each i ∈ {1, 2, ..., N}, n i One observation value.

[0102] More generally:

[0103] y i =X i β i +Z i γ i +ε i i = 1, 2, ..., N

[0104] It can be written as:

[0105] y i =f(X1,X2,...,X) p )+Zi γ i +ε i i = 1, 2, ..., N

[0106] Where, f(X1,X2,...,X) p The fixed effects component used to explain this can be linear or nonlinear. i γ i This represents the random utility component. This modeling formula greatly improves the adaptability and flexibility of the model.

[0107] Analysis of results from a linear random effects mixture model:

[0108] By performing regression analysis on customer credit card data, we can obtain estimated parameters, variance, equation significance test, and parameter significance verification. We can also predict the total credit card spending Y of N customers. This avoids the situation in related technologies where the same regression equation is used, which can only accurately output the total credit card spending Y of the customer as a whole, but cannot distinguish the differences between individual customers, resulting in inaccurate output of the total credit card spending of each customer.

[0109] Optionally, in the financial data processing method provided in this application embodiment, N mixed models are determined based on the fitting results, including: verifying each fitting result based on a target verification strategy to obtain the verification result of each linear model, wherein the target verification strategy includes at least one of the following: significance test, regression coefficient test and goodness-of-fit test; if the verification results of S linear models all meet the preset conditions, N mixed models are determined based on the fitting results of the random slope intercept model.

[0110] For example, (1) using a general linear model: y=Xβ+ε, i=1,2,...,N, to fit the dataset (i.e. the second dataset), the maximum likelihood method can be used to obtain the estimated parameter β and the variance of the model. Statistical tests include the significance test of the regression equation, the regression coefficient test and the goodness-of-fit test.

[0111] (2) If the validation results of a general linear model meet the preset conditions, a stochastic intercept model can be used: y i =X i β i +γ i +ε i The dataset is refitted to obtain estimated parameters β and γ, variance, and statistical test statistic. The preset condition can refer to the accuracy of the prediction results of the fitted model for predicting consumption data reaching a preset threshold, for example, the variance is less than a preset variance threshold.

[0112] (3) If the validation results of the random intercept model meet the preset conditions, the random slope model can be used: y i =X i β i +ε i The dataset is refitted, the parameters are estimated, and the estimated parameters β and γ, variance, and statistical test statistic are obtained.

[0113] (4) If the verification results of the random slope model meet the preset conditions, the random intercept slope model is adopted: y i =X i β i +Z i γ i +ε i For each i = 1, 2, ..., N, fit the dataset to estimate the parameters β and γ, variance, and statistical test result. Then, add random vector variables X2 and X3 as random effects to the model, update the model, and re-estimate the parameters β and γ, variance, and statistical test result.

[0114] In one alternative example, the model with the most significant regression equation and the smallest variance can be selected based on the estimated parameters, variance, and statistical tests of the regression equation, regression coefficient tests, and goodness-of-fit tests of each model. This achieves the technical effect of improving the accuracy of the output results of the mixed model.

[0115] Optionally, in the financial data processing method provided in the embodiments of this application, after determining N mixed models based on the fitting results of the random slope intercept model, the method further includes: extracting the random slope and random intercept of each of the N mixed models to obtain a third dataset; and constructing a second prediction model based on the third dataset, wherein the second prediction model includes: a model with random slope and random intercept as dependent variables and financial data generated by the target financial account as independent variables.

[0116] In this embodiment, the intercept and slope can also be used as dependent variables to fit a second prediction model. The random intercept and slope allow for differences in levels (intercepts) and associations (slopes) between individuals. For a specific individual (unit), even if repeated measures data may exist, there is only one intercept and one slope. Therefore, although the random intercept and slope are derived from repeated test data (individual level), the intercept and slope themselves are still repeated measures data at the population level. The intercept or slope for each individual does not change with the estimation. Assuming that the intercept and slope do indeed vary from individual to individual, the predicted values ​​of the intercept and slope can be determined. To be more specific, Figure 3 This is based on the individual and group relationship diagram provided in the embodiments of this application, such as... Figure 3As shown. The dependent variable y is predicted from the individual layer x using random slope and random intercept. The random intercept and random slope are themselves outcome variables of the group layer, and can also be predicted by the predictors of the group layer. Therefore, the observed total credit card spending Y can be predicted from the random intercept and random slope of the independent variables X2 and X3. In this embodiment, for customers whose independent variables X2 and X3 are higher than the average, their total credit card spending Y is also higher (using X2 and X3 to predict the random intercept); or for customers whose independent variables X2 and X3 are lower than the average, their total credit card spending Y changes less (the random slope is relatively flat). Since the average independent variables X2 and X3 do not change, they are inter-individual variables. Figure 3 It explains the relationship between them.

[0117] It should be noted that in this embodiment, the random intercept can represent the minimum value of a customer's total credit card spending, and the random slope can represent the magnitude of the influence of the independent variables X1, X2, X3, X4, X5, and X6 on the growth of each customer's total credit card spending.

[0118] Suppose there are 100 employees. We can observe whether customers with X2 and X3 who are more than the average also have higher total credit card spending.

[0119] For customers whose independent variables X2 and X3 are less than the average, is it also true that their total credit card spending Y changes less (the random slope is relatively flat)?

[0120] The fixed intercept and fixed slope of the total credit card spending Y can both be predicted by the predictors (X2, X3, X4, X5, X6) of the group stratum.

[0121] In this embodiment, the total credit card spending Y can be predicted using the individual and group layers of recorded dates, and the random intercept and slope can be used to predict the random intercept and slope using the average number of transactions X2 and X3.

[0122] If the random intercept γ of the random parameters for each customer N0 and random slope γ N2 and γ N3 This is the third dataset, as shown in Table 1.

[0123] Table 1

[0124]

[0125]

[0126] With γ i0 Establish a coordinate system with x as the x-axis and yi as the y-axis, and obtain Figure 4 ,in, Figure 4This is a schematic diagram of the first coordinate system provided according to an embodiment of this application;

[0127] With γ i2 Establish a coordinate system with x as the x-axis and y as the y-axis, and obtain... Figure 5 ,in, Figure 5 This is a schematic diagram of the second coordinate system provided according to an embodiment of this application.

[0128] from Figure 4 and Figure 5 It can be determined that with a sufficiently large customer sample, the random intercept (work unit of strangers) and random slope (job level of strangers) can not only be calculated using the least squares method, but the random intercept can also be predicted based on the dependent variable total credit card spending Y, and the random slope can be predicted based on the independent variables X2, X3 and the dependent variable total credit card spending Y. (Independent variables X1, X2, X3, X, X5, X6, dependent variable Y) The resulting second prediction model is:

[0129] γ i0 =c0+c1y i +ε 0i

[0130] γ i2 =d0+d1y i +ε 2i i = 1, 2, ..., N

[0131] γ i3 =f0+f1y i +ε 3i

[0132] Where c0, c1, d0, d1, f0, and f1 are the estimated parameters of the second prediction model, and in the second prediction model, yi is the independent variable, and the random intercept γ is the random intercept γ. i0 and random slope γ i2 and γ i3 As the dependent variable, the random intercept of the total credit card spending Y and the fixed slope of the total credit card spending Y relative to X1, X2, X3, X4, X5 and X6 can be predicted by the predictors of the population layer (average independent variables X1, X2, X3, X4, X5 and X6).

[0133] In this embodiment, a third prediction model can be constructed based on the random slope and random intercept of each of the N mixture models. Using the third prediction model, given the total transaction amount of a target financial account of an object in a target time period, the random intercept and random slope of the associated mixture model of that object can be predicted. The random intercept can represent the minimum value of the total consumption of the target financial account of the object, and the random slope can represent the magnitude of the influence of the random effects of the independent variables X2 and X3 on the growth of the total consumption of the target financial account of the object.

[0134] Optionally, in the financial data processing method provided in the embodiments of this application, the financial data of the target financial account includes: the total transaction amount of the target financial account in the target time period. After constructing the second prediction model based on the third dataset, the method further includes: obtaining the financial data of the target financial account of the third object in the financial institution to obtain the second financial data; inputting the second financial data into the second prediction model and outputting the target slope and the target intercept, wherein the target slope is used to determine the degree of influence of the second financial data on the attribute data of the third object or the purchase of financial products by the third object, and the target intercept is used to determine the minimum total transaction amount of the third object at multiple time points in the target time period.

[0135] In this embodiment, given the total transaction amount (e.g., total consumption amount) of a target financial account of a certain object (corresponding to the third object) during a target time period, the total transaction amount of the object during the target time period can be input into the third prediction model. The third prediction model outputs the random intercept and random slope of the associated mixture model of the object. The random intercept can represent the minimum value of the total consumption amount of the target financial account of the object, and the random slope can represent the magnitude of the influence of the random effects of the independent variables X2 and X3 on the growth of the total consumption amount of the target financial account of the object. Furthermore, the random effects corresponding to the random effects of the object can be inferred based on the random slope. For example, if the random effects associated with the mixture model corresponding to the random slope are working city A and occupation B, then the working city A and occupation B of the object can be deduced, thus achieving the purpose of predicting customer information based on the total consumption amount.

[0136] Optionally, in the financial data processing method provided in the embodiments of this application, after inputting the first dataset into the first prediction model and outputting the first financial data, the method further includes: obtaining a preset mapping rule, matching the first financial data with the preset mapping rule to obtain a second matching result, wherein the preset mapping rule includes: the mapping relationship between the financial data of the target financial account and T types of financial products, where T is a positive integer; and determining the target financial product based on the second matching result, wherein the target financial product includes: the financial product to be recommended to the first object.

[0137] In an optional example, the pre-defined mapping rules may include: mapping rules between the financial data of the target financial account and T types of financial products. For example, mapping rules between the total transaction amount of the target financial account and T types of financial products. Specifically, the total transaction amount within different pre-defined amount ranges may have a mapping relationship with different financial products.

[0138] In this embodiment, the total transaction amount (total consumption data) involved in the first financial data can be matched with a preset mapping rule. Based on the preset amount range to which the total transaction amount belongs, the financial product associated with the total transaction amount (i.e., the target financial product) can be determined. Then, the target financial product can be recommended to the first object. By recommending financial products to the first object based on the total transaction amount of the first object, the situation of recommending financial products that the first object does not have purchasing power is avoided, thereby achieving the technical effect of improving the accuracy of financial product recommendation.

[0139] It should be noted that the stochastic intercept slope model is used: y i =X i β i +Z i γ i +ε i For i = 1, 2, ..., N, when fitting the dataset and estimating the parameters to obtain the estimated parameters β and γ, if Y is a variable of 0 or 1, or a fixed variable of multiple terms (0, 1, 2, 3, 4, 5), the stochastic intercept slope model can also be written as a logistic model of generalized linear stochastic intercept slope (logistic regression model):

[0140]

[0141]

[0142] More generally, y i =X i β i +Z i γ i +ε i For i = 1, 2, ..., N, it can also be written as:

[0143] y i =f(X1,X2,...,X) p )+Z i γ i +ε i i = 1, 2, ..., N

[0144] Where, f(X1,X2,...,X) p The fixed effects component used to explain this can be linear or nonlinear. i γi This is the random utility component. The dependent variable, the observer Y, can be a continuously distributed variable that is normally distributed and fitted with a linear model, or it can be a binomial distribution that is fitted with a logistic model, or it can be an exponential distribution, a Gamma distribution, an inverse Gaussian distribution, or a Poisson distribution. This modeling formula greatly improves the adaptability and flexibility of the model. The modeling formulas for various distributions are shown in Table 2.

[0145] Table 2

[0146]

[0147] In this embodiment, where customer information is used with the customer's knowledge and authorization for collection and use, the accuracy is higher than that of a general regression analysis model. A stochastic intercept and slope model is used, with independent variables X1, X2, X3, X4, X5, and X6 for N customers. Each customer has ni observations: total credit card spending Y and independent variables X2 and X3. N regression models with fixed and random effects are established for the N customers. The fixed effect parameters are the same, revealing characteristics of the customer group; the random effect parameters are different, revealing individual characteristics. A general regression model only establishes a fixed-effect regression model for the customer group, only revealing characteristics of the customer group; or it establishes N different regression models for N customers, repeatedly calculating the regression model parameters N times, only revealing individual customer characteristics. A general regression model cannot simultaneously reveal characteristics of both the customer group and individuals. It utilizes historical credit card data more efficiently than a time series model. A stochastic intercept and stochastic slope model is used, with N customers, each with ni observations, and these ni different observations are recorded at different times t. The number of observations, observation intervals, and the number of measurements may vary for each client. Compared to time series models, which require a large number of observations at equal intervals, the model used in this embodiment can effectively utilize data with unequal observation times.

[0148] In this embodiment, the linear stochastic intercept and stochastic slope model can be extended to the logistic model of the generalized linear stochastic intercept and slope, and further extended to the general form y i =f(X1,X2,...,X) p )+Z i γ i +ε i In the formula, f(X1,X2,...,X) p The fixed effects component used to explain this can be linear or nonlinear. i γ iThis is the random utility component. The dependent variable, the observer Y, can be a normally distributed continuous variable fitted with a linear model, a binomial variable fitted with a logistic model, or an exponential, Gamma, inverse Gaussian, or Poisson distribution. This greatly improves the model's adaptability and flexibility.

[0149] It should be noted that the steps shown in the flowchart in the accompanying drawings can be executed in a computer system such as a set of computer-executable instructions, and although a logical order is shown in the flowchart, in some cases the steps shown or described may be executed in a different order than that shown here.

[0150] Example 2

[0151] This application also provides a financial data processing apparatus. It should be noted that the financial data processing apparatus of this application can be used to execute the financial data processing method provided in this application. The following describes the financial data processing apparatus provided in this application.

[0152] According to embodiments of this application, an apparatus for implementing the above-described financial data processing method is also provided, such as... Figure 6 As shown, the device includes: a first acquisition unit 61, a screening unit 62, and a first processing unit 63.

[0153] The first acquisition unit 61 is used to acquire the first dataset of the first object, wherein the first dataset includes: attribute data of the first object and M types of financial data generated by the first object in purchasing financial products during a historical period, where M is a positive integer;

[0154] The filtering unit 62 is used to filter the first prediction model from N mixture models based on the first dataset. The model types of the N mixture models include: random slope intercept model, which includes: model built based on random slope and random intercept, where N is a positive integer.

[0155] The first processing unit 63 is used to input the first dataset into the first prediction model and output the first financial data, wherein the first financial data includes: financial data related to the target financial account of the first object in the financial institution, and the financial data includes: consumption data.

[0156] The financial data processing apparatus provided in this application embodiment includes a first acquisition unit 61, used to acquire a first dataset of a first object, wherein the first dataset includes: attribute data of the first object and M types of financial data generated by the first object purchasing financial products during a historical period, where M is a positive integer; a filtering unit 62, used to filter a first prediction model from N mixed models based on the first dataset, wherein the model types of the N mixed models include: a stochastic slope intercept model, and the stochastic slope intercept model includes: a model constructed based on a stochastic slope and a stochastic intercept, where N is a positive integer; and a first processing unit 63, used to input the first dataset into the first prediction model and output first financial data, wherein the first financial data includes: financial data related to the target financial account of the first object in a financial institution, and the financial data includes: consumption data. This solves the technical problem of low accuracy in related technologies that use a single regression equation to output consumption data of all customers' financial accounts.

[0157] In this embodiment, based on the dataset of the first object, a first prediction model associated with the first object is selected from N hybrid models. The consumption data of the target financial account of the first object is output using the first prediction model. This avoids the situation in related technologies where a single regression equation is used to output the consumption data of the target financial accounts of all customers, which cannot distinguish individual differences. This achieves the technical effect of accurately determining the consumption data of the financial account of a single customer.

[0158] Furthermore, in the financial data processing apparatus provided in this application embodiment, the filtering unit includes: an acquisition subunit, configured to acquire N second datasets, wherein the N second datasets include: attribute data of each of the N second objects and M types of financial data generated by the purchase of financial products by the second object; a matching subunit, configured to match the first dataset and the N second datasets to obtain a first matching result, wherein the first matching result is used to indicate the second dataset among the N second datasets that successfully matches the first dataset; and a first determination subunit, configured to determine the hybrid model associated with the second dataset indicated by the matching result as a first prediction model.

[0159] Furthermore, in the financial data processing apparatus provided in this application embodiment, N hybrid models are obtained through the following sub-units: a second determining sub-unit, used to determine S linear models, wherein the model types of the S linear models include at least one of the following: a stochastic slope model, a stochastic intercept model, and a stochastic slope-intercept model, where S is a positive integer; and a fitting sub-unit, used to perform data fitting on N second datasets using each linear model to obtain the fitting result of each linear model, and based on the fitting result, determine N hybrid models, wherein the data fitting is used to determine the parameter values ​​of the unknown parameters in the linear model.

[0160] Furthermore, in the financial data processing apparatus provided in this application embodiment, the fitting subunit includes: a verification module, used to verify each fitting result based on a target verification strategy to obtain a verification result for each linear model, wherein the target verification strategy includes at least one of the following: significance test, regression coefficient test, and goodness-of-fit test; and a determination module, used to determine N mixed models based on the fitting result of the random slope intercept model when the verification results of the S linear models all meet the preset conditions.

[0161] Furthermore, in the financial data processing apparatus provided in the embodiments of this application, the financial data processing apparatus further includes: an extraction unit, used to extract the random slope and random intercept of each of the N mixture models after determining N mixture models based on the fitting results of the random slope intercept model, to obtain a third dataset; and a construction subunit, used to construct a second prediction model based on the third dataset, wherein the second prediction model includes: a model with random slope and random intercept as dependent variables and financial data generated by the target financial account as independent variables.

[0162] Furthermore, in the financial data processing apparatus provided in this application embodiment, the financial data of the target financial account includes: the total transaction amount of the target financial account during the target time period. The financial data processing apparatus further includes: a second acquisition unit, used to acquire the financial data of the target financial account of the third object in the financial institution after constructing a second prediction model based on the third dataset, to obtain the second financial data; and a second processing unit, used to input the second financial data into the second prediction model and output a target slope and a target intercept, wherein the target slope is used to determine the degree of influence of the second financial data on the attribute data of the third object or the purchase of financial products by the third object, and the target intercept is used to determine the minimum total transaction amount of the third object at multiple time points during the target time period.

[0163] Furthermore, in the financial data processing apparatus provided in the embodiments of this application, the financial data processing apparatus further includes: a third acquisition unit, configured to, after inputting the first dataset into the first prediction model and outputting the first financial data, acquire a preset mapping rule, match the first financial data and the preset mapping rule to obtain a second matching result, wherein the preset mapping rule includes: a mapping relationship between the financial data of the target financial account and T types of financial products, where T is a positive integer; and a determination unit, configured to, based on the second matching result, determine the target financial product, wherein the target financial product includes: a financial product to be recommended to the first object.

[0164] It should be noted that the first acquisition unit 61, the filtering unit 62, and the first processing unit 63 mentioned above correspond to steps S201 to S203 in Embodiment 1. The two modules and the corresponding steps implement the same instances and application scenarios, but are not limited to the content disclosed in Embodiment 1. It should be noted that the above modules or units can be hardware or software components stored in memory (e.g., memory 104) and processed by one or more processors (e.g., processors 102a, 102b, ..., 102n). The above modules can also be part of a device and run in the computer terminal 10 provided in Embodiment 1.

[0165] Example 3

[0166] Embodiments of this application may provide an electronic device. Figure 7 This is a structural block diagram of an electronic device according to an embodiment of this application. Figure 7 As shown, the electronic device may include: one or more ( Figure 7 Only one of the following is shown: processor 702, memory 704, memory controller, and peripheral interface, wherein the peripheral interface is connected to the radio frequency module, audio module, and display.

[0167] The memory can be used to store software programs and modules, such as the program instructions / modules corresponding to the methods and apparatus in the embodiments of this application. The processor executes various functional applications and data processing by running the software programs and modules stored in the memory, thereby implementing the above-described methods. The memory may include high-speed random access memory, and may also include non-volatile memory, such as one or more magnetic storage devices, flash memory, or other non-volatile solid-state memory. In some instances, the memory may further include memory remotely located relative to the processor, and these remote memories can be connected to the terminal via a network. Examples of such networks include, but are not limited to, the Internet, corporate intranets, local area networks, mobile communication networks, and combinations thereof.

[0168] The processor can invoke information and applications stored in the memory via a transmission device to perform the following steps: Obtain a first dataset for a first object, wherein the first dataset includes: attribute data of the first object and M types of financial data generated by the first object's purchase of financial products during a historical time period, where M is a positive integer; Based on the first dataset, select a first prediction model from N mixed models, wherein the model types of the N mixed models include: a stochastic slope intercept model, and the stochastic slope intercept model includes: a model constructed based on a stochastic slope and a stochastic intercept, where N is a positive integer; Input the first dataset into the first prediction model and output first financial data, wherein the first financial data includes: financial data related to the first object's target financial account in a financial institution, and the financial data includes: consumption data.

[0169] The processor can also invoke information and applications stored in the memory via a transmission device to perform the following steps: based on a first dataset, select a first prediction model from N hybrid models, including: obtaining N second datasets, wherein the N second datasets include: attribute data of each of the N second objects and M types of financial data generated by the purchase of financial products by the second object; matching the first dataset and the N second datasets to obtain a first matching result, wherein the first matching result is used to indicate the second dataset among the N second datasets that successfully matches the first dataset; and determining the hybrid model associated with the second dataset indicated by the matching result as the first prediction model.

[0170] The processor can also call the information and application programs stored in the memory through the transmission device to perform the following steps: N mixture models are obtained by: determining S linear models, wherein the model types of the S linear models include at least one of the following: stochastic slope model, stochastic intercept model, and stochastic slope-intercept model, where S is a positive integer; using each linear model to fit data to N second datasets to obtain the fitting result of each linear model, and based on the fitting result, determining N mixture models, wherein the data fitting is used to determine the parameter values ​​of the unknown parameters in the linear model.

[0171] The processor can also call the information and application stored in the memory through the transmission device to perform the following steps: Based on the fitting results, determine N mixture models, including: verifying each fitting result based on the target verification strategy to obtain the verification result of each linear model, wherein the target verification strategy includes at least one of the following: significance test, regression coefficient test and goodness of fit test; if the verification results of S linear models all meet the preset conditions, determine N mixture models based on the fitting results of the random slope intercept model.

[0172] The processor can also call the information and application stored in the memory through the transmission device to perform the following steps: after determining N mixture models based on the fitting results of the random slope intercept model, the processor further includes: extracting the random slope and random intercept of each mixture model from the N mixture models to obtain a third dataset; and constructing a second prediction model based on the third dataset, wherein the second prediction model includes: a model with random slope and random intercept as dependent variables and financial data generated by the target financial account as independent variables.

[0173] The processor can also invoke information and applications stored in the memory via a transmission device to perform the following steps: The financial data of the target financial account includes: the total transaction amount of the target financial account in the target time period. After constructing the second prediction model based on the third dataset, the process further includes: obtaining the financial data of the target financial account of the third object in the financial institution to obtain the second financial data; inputting the second financial data into the second prediction model and outputting the target slope and the target intercept, wherein the target slope is used to determine the degree of influence of the second financial data on the attribute data of the third object or the purchase of financial products by the third object, and the target intercept is used to determine the minimum total transaction amount of the third object at multiple time points in the target time period.

[0174] The processor can also call the information and application stored in the memory through the transmission device to perform the following steps: after inputting the first dataset into the first prediction model and outputting the first financial data, it further includes: obtaining a preset mapping rule, matching the first financial data and the preset mapping rule to obtain a second matching result, wherein the preset mapping rule includes: the mapping relationship between the financial data of the target financial account and T types of financial products, where T is a positive integer; based on the second matching result, determining the target financial product, wherein the target financial product includes: the financial product to be recommended to the first object.

[0175] This application provides a financial data processing solution. Based on a dataset of a first object, a first prediction model associated with the first object is selected from N hybrid models. This first prediction model is then used to output the consumption data of the target financial account of the first object. This avoids the problem in related technologies where a single regression equation is used to output the consumption data of all customers' target financial accounts, failing to distinguish individual differences. Therefore, this achieves the technical effect of accurately determining the consumption data of a single customer's financial account. Furthermore, it solves the technical problem of low accuracy in related technologies that use a single regression equation to output the consumption data of all customers' financial accounts.

[0176] Those skilled in the art will understand that Figure 7 The structure shown is for illustrative purposes only. Electronic devices can also be smartphones, tablets, handheld computers, mobile internet devices (MIDs), PADs, and other terminal devices. Figure 7 This does not limit the structure of the aforementioned electronic device. For example, electronic devices may also include components that are more... Figure 7 The more or fewer components shown (such as network interfaces, display devices, etc.), or having the same Figure 7 The different configurations shown.

[0177] Those skilled in the art will understand that all or part of the steps in the various methods of the above embodiments can be implemented by a program instructing the hardware related to the terminal device. The program can be stored in a computer-readable storage medium, which may include: flash drive, read-only memory (ROM), random access memory (RAM), disk or optical disk, etc.

[0178] Example 4

[0179] Embodiments of this application also provide a storage medium. Optionally, in this embodiment, the storage medium can be used to store the program code executed by the financial data processing method provided in Embodiment 1.

[0180] Optionally, in this embodiment, the storage medium may be located in any computer terminal in a group of computer terminals in a computer network, or in any mobile terminal in a group of mobile terminals.

[0181] This application also provides a computer program product, which, when executed on a data processing device, is suitable for performing steps of a financial data processing method.

[0182] The serial numbers of the above embodiments of the present application are for description only and do not represent the advantages or disadvantages of the embodiments.

[0183] In the above embodiments of this application, the descriptions of each embodiment have different focuses. For parts not described in detail in a certain embodiment, please refer to the relevant descriptions of other embodiments.

[0184] In the several embodiments provided in this application, it should be understood that the disclosed technical content can be implemented in other ways. The device embodiments described above are merely illustrative; for example, the division of units is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the displayed or discussed mutual coupling, direct coupling, or communication connection may be through some interfaces; the indirect coupling or communication connection between units or modules may be electrical or other forms.

[0185] The units described as separate components may or may not be physically separate, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed across multiple network units. Some or all of these units may be selected to achieve the purpose of this embodiment according to actual needs.

[0186] Furthermore, the functional units in the various embodiments of this application can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or as a software functional unit.

[0187] If the integrated unit is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of this application. The aforementioned storage medium includes various media capable of storing program code, such as a USB flash drive, read-only memory (ROM), random access memory (RAM), portable hard drive, magnetic disk, or optical disk.

[0188] The above is only 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 of 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 processing financial data, characterized in that, include: Obtain the first dataset of the first object, wherein the first dataset includes: attribute data of the first object and M types of financial data generated by the first object purchasing financial products during a historical time period, where M is a positive integer; Based on the first dataset, a first prediction model is selected from N hybrid models, wherein the model types of the N hybrid models include: random slope intercept models, wherein the random slope intercept models include: models constructed based on random slope and random intercept, and N is a positive integer; Based on the first dataset, a first predictive model is selected from N hybrid models, including: obtaining N second datasets, wherein the N second datasets include: attribute data of each of the N second objects and M types of financial data generated by the purchase of financial products by the second object; matching the first dataset and the N second datasets to obtain a first matching result, wherein the first matching result is used to indicate the second dataset among the N second datasets that successfully matches the first dataset; and determining the hybrid model associated with the second dataset indicated by the matching result as the first predictive model. The N hybrid models are obtained by: determining S linear models, wherein the model types of the S linear models include at least one of the following: a stochastic slope model, a stochastic intercept model, and a stochastic slope-intercept model, where S is a positive integer; performing data fitting on the N second datasets using each of the linear models to obtain the fitting result of each linear model, and determining the N hybrid models based on the fitting result, wherein the data fitting is used to determine the parameter values ​​of the unknown parameters in the linear model. Based on the fitting results, N mixed models are determined, including: verifying each fitting result based on a target verification strategy to obtain a verification result for each linear model, wherein the target verification strategy includes at least one of the following: significance test, regression coefficient test, and goodness-of-fit test; if the verification results of S linear models all meet the preset conditions, N mixed models are determined based on the fitting results of the random slope intercept model. The first dataset is input into the first prediction model, and the first financial data is output. The first financial data includes: financial data related to the target financial account of the first object in a financial institution, and the financial data includes: consumption data.

2. The processing method according to claim 1, characterized in that, After determining N hybrid models based on the fitting results of the random slope intercept model, the process further includes: Extract the random slope and random intercept of each of the N mixture models to obtain the third dataset; Based on the third dataset, a second prediction model is constructed, wherein the second prediction model includes a model with random slope and random intercept as dependent variables and financial data generated by the target financial account as independent variables.

3. The processing method according to claim 2, characterized in that, The financial data of the target financial account includes: the total transaction amount of the target financial account during the target time period, and after constructing the second prediction model based on the third dataset, it also includes: Obtain the financial data of the target financial account of the third object in the financial institution to obtain the second financial data; The second financial data is input into the second prediction model, and the target slope and target intercept are output. The target slope is used to determine the degree of influence of the second financial data on the attribute data of the third object or the purchase of financial products by the third object. The target intercept is used to determine the minimum total transaction amount of the third object at multiple time points in the target time period.

4. The processing method according to claim 1, characterized in that, After inputting the first dataset into the first prediction model and outputting the first financial data, the process further includes: Obtain a preset mapping rule, match the first financial data with the preset mapping rule to obtain a second matching result, wherein the preset mapping rule includes: the mapping relationship between the financial data of the target financial account and T types of financial products, where T is a positive integer; Based on the second matching result, a target financial product is determined, wherein the target financial product includes: a financial product to be recommended to the first object.

5. A financial data processing device, characterized in that, include: The first acquisition unit is used to acquire the first dataset of the first object, wherein the first dataset includes: attribute data of the first object and M types of financial data generated by the first object in purchasing financial products during a historical time period, where M is a positive integer; A filtering unit is used to filter out a first prediction model from N hybrid models based on the first dataset, wherein the model types of the N hybrid models include: random slope intercept models, wherein the random slope intercept models include: models constructed based on random slope and random intercept, and N is a positive integer; The filtering unit includes: an acquisition subunit for acquiring N second datasets, wherein the N second datasets include: attribute data of each of the N second objects and M types of financial data generated by the purchase of financial products by the second object; a matching subunit for matching the first dataset and the N second datasets to obtain a first matching result, wherein the first matching result is used to indicate the second dataset among the N second datasets that successfully matches the first dataset; and a first determination subunit for determining the hybrid model associated with the second dataset indicated by the matching result as the first prediction model. The N hybrid models are obtained through the following sub-units: a second determining sub-unit, used to determine S linear models, wherein the model types of the S linear models include at least one of the following: a stochastic slope model, a stochastic intercept model, and a stochastic slope-intercept model, where S is a positive integer; and a fitting sub-unit, used to perform data fitting on the N second datasets using each of the linear models to obtain the fitting result of each linear model, and based on the fitting result, to determine the N hybrid models, wherein the data fitting is used to determine the parameter values ​​of the unknown parameters in the linear model. The fitting subunit includes: a verification module, used to verify each fitting result based on a target verification strategy to obtain a verification result for each linear model, wherein the target verification strategy includes at least one of the following: significance test, regression coefficient test and goodness-of-fit test; and a determination module, used to determine N mixed models based on the fitting result of the random slope intercept model when the verification results of the S linear models all meet the preset conditions. The first processing unit is configured to input the first dataset into the first prediction model and output the first financial data, wherein the first financial data includes: financial data related to the target financial account of the first object in a financial institution, and the financial data includes: consumption data.

6. A computer-readable storage medium, characterized in that, The computer-readable storage medium includes a stored executable program, wherein, when the executable program is executed, it controls the device on which the computer-readable storage medium is located to perform the financial data processing method according to any one of claims 1 to 4.

7. An electronic device, characterized in that, include: Memory, which stores executable programs; A processor for running the program, wherein the program, when running, performs the method for processing financial data according to any one of claims 1 to 4.

8. A computer program product comprising computer instructions, characterized in that, When the computer instructions are executed by the processor, they implement the steps of the financial data processing method according to any one of claims 1 to 4.

Images