CN112116169B - User behavior determining method and device and electronic equipment - Google Patents
User behavior determining method and device and electronic equipment Download PDFInfo
- Publication number
- CN112116169B CN112116169B CN202011048577.1A CN202011048577A CN112116169B CN 112116169 B CN112116169 B CN 112116169B CN 202011048577 A CN202011048577 A CN 202011048577A CN 112116169 B CN112116169 B CN 112116169B
- Authority
- CN
- China
- Prior art keywords
- amount
- target financial
- project
- target
- sample
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q10/00—Administration; Management
- G06Q10/04—Forecasting or optimisation specially adapted for administrative or management purposes, e.g. linear programming or "cutting stock problem"
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/045—Combinations of networks
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
- G06N3/084—Backpropagation, e.g. using gradient descent
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/12—Computing arrangements based on biological models using genetic models
- G06N3/126—Evolutionary algorithms, e.g. genetic algorithms or genetic programming
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q40/00—Finance; Insurance; Tax strategies; Processing of corporate or income taxes
- G06Q40/02—Banking, e.g. interest calculation or account maintenance
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Business, Economics & Management (AREA)
- Biophysics (AREA)
- Life Sciences & Earth Sciences (AREA)
- Health & Medical Sciences (AREA)
- General Physics & Mathematics (AREA)
- Artificial Intelligence (AREA)
- Computational Linguistics (AREA)
- Computing Systems (AREA)
- Software Systems (AREA)
- Molecular Biology (AREA)
- General Health & Medical Sciences (AREA)
- Strategic Management (AREA)
- Evolutionary Computation (AREA)
- Data Mining & Analysis (AREA)
- Biomedical Technology (AREA)
- Mathematical Physics (AREA)
- General Engineering & Computer Science (AREA)
- Economics (AREA)
- Evolutionary Biology (AREA)
- Finance (AREA)
- Development Economics (AREA)
- Marketing (AREA)
- Accounting & Taxation (AREA)
- Human Resources & Organizations (AREA)
- General Business, Economics & Management (AREA)
- Bioinformatics & Computational Biology (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Operations Research (AREA)
- Technology Law (AREA)
- Tourism & Hospitality (AREA)
- Quality & Reliability (AREA)
- Entrepreneurship & Innovation (AREA)
- Game Theory and Decision Science (AREA)
- Physiology (AREA)
- Genetics & Genomics (AREA)
- Financial Or Insurance-Related Operations Such As Payment And Settlement (AREA)
Abstract
The application discloses a method and a device for determining user behaviors and electronic equipment, wherein the method comprises the following steps: receiving a project handling request of a target user, wherein the project handling request corresponds to a target financial project to be handled, and the project handling request at least comprises handling amount; obtaining project data of a target financial project, wherein the project data at least comprises: the method comprises the steps of presetting total amount of target financial items, first amount of full money handling target financial items, second amount of pre-purchase and transaction target financial items, first user number of full money handling target financial items and second user number of pre-purchase and transaction target financial items; inputting project data into a pre-trained neural network model to obtain a predicted residual amount of a target financial project output by the neural network model; and obtaining a behavior determination result at least according to the predicted remaining amount and the transaction amount, wherein the behavior determination result characterizes whether the target user transacts the target financial project.
Description
Technical Field
The present application relates to the field of financial wind control technologies, and in particular, to a method and an apparatus for determining user behavior, and an electronic device.
Background
Currently, financial institutions such as banks and the like typically offer users a plurality of financial products, and each financial product is typically limited by a total amount of items, and after exceeding the total amount of items, the financial institutions no longer recommend and sell the financial product to the customer. In order to facilitate customers, financial institutions typically provide customers with pre-purchase services, namely: the customer may pay a subscription to indicate a desire to purchase a financial product, and the customer may default within a certain period of time, at which time the financial institution refunds the subscription to the customer. And after the customer violates the contract, the previously pre-purchased amount occupied in the total amount of the item is released.
However, before the customer breaks the contract, the financial institution cannot determine whether to continue to transact the application of purchasing the financial product for other customers, if the financial institution continues to transact the application of purchasing the financial product for other customers but no customer breaks the contract or the amount released by the customer breaks the contract is low, the total amount of the project is excessively increased; if the financial institution does not continue to transact the application of purchasing the financial product for other customers but there is a customer breach and the amount released by the customer breach is high, the amount released in the financial product may not be occupied by any customer, resulting in a waste of the amount.
Therefore, there is a need for a solution that can determine whether to continue to transact financial products with other customers.
Disclosure of Invention
In view of the above, the present application provides a method and apparatus for determining user behavior, and an electronic device, as follows:
a method of determining user behavior, comprising:
receiving a project handling request of a target user, wherein the project handling request corresponds to a target financial project to be handled, and the project handling request at least comprises handling amount;
obtaining project data of the target financial project, wherein the project data at least comprises: the target financial item comprises a preset total amount, a first amount for transacting the target financial item, a second amount for pre-purchasing the target financial item, a first user number for transacting the target financial item and a second user number for pre-purchasing the target financial item;
inputting the project data into a pre-trained neural network model to obtain the predicted residual amount of the target financial project output by the neural network model;
the neural network model is trained by using a plurality of first sample data with residual credit labels, wherein the first sample data comprises a sample total credit of the target financial project, a sample amount of the target financial project, a sample user number of the target financial project and the sample user number of the target financial project;
And obtaining a behavior determination result at least according to the predicted remaining amount and the transacted amount, wherein the behavior determination result characterizes whether the target user transacts the target financial project.
In the above method, preferably, obtaining a behavior determination result at least according to the predicted remaining sum and the transaction amount includes:
inputting the predicted residual amount, the preset total amount and the transacted amount into a pre-trained classification model to obtain a behavior determination result output by the classification model; the behavior determination result characterizes whether the target user transacts the target financial project;
the classification model is obtained by training a plurality of second sample data with transaction tags, the second sample data comprises the sample total amount, the sample residual amount and the sample transaction amount, and the transaction tags represent whether to transact a target financial item corresponding to the sample transaction amount.
In the above method, preferably, obtaining a behavior determination result at least according to the predicted remaining sum and the transaction amount includes:
comparing the transacted amount, the predicted remaining amount and the preset total amount to obtain a comparison result;
And obtaining a behavior determination result according to the comparison result, wherein the behavior determination result characterizes whether the target user handles the target financial project.
In the above method, preferably, the behavior determination result is characterized as that the target user handles the target financial item when the comparison result is that the ratio of the handled amount to the preset total amount is smaller than a ratio threshold.
In the above method, preferably, the behavior determination result is characterized as that the target user handles the target financial item when the comparison result indicates that the ratio of the handled amount to the preset total amount is greater than or equal to the ratio threshold and the handled amount is less than or equal to the predicted remaining amount.
In the above method, preferably, when the comparison result indicates that the ratio of the transacted amount to the preset total amount is greater than or equal to the ratio threshold, the transacted amount is greater than the predicted remaining amount, and the difference between the transacted amount and the predicted remaining amount is less than or equal to the difference threshold, the behavior determination result indicates that the target user transacts the target financial item;
And under the condition that the comparison result represents that the ratio of the transacted amount to the preset total amount is larger than or equal to the ratio threshold, the transacted amount is larger than the predicted remaining amount, and the difference between the transacted amount and the predicted remaining amount is larger than the difference threshold, the behavior determination result represents that the target financial project is not transacted for the target user.
The above method, preferably, further comprises:
obtaining third sample data, the third sample data comprising: the sample total amount of the target financial item, the sample user number of the target financial item, and the sample user number of the target financial item;
testing the neural network model by using the third sample data to obtain a test result of the neural network model; the test results at least characterize the accuracy of the neural network model.
The above method, preferably, further comprises:
and outputting the predicted residual amount of the target financial project to prompt the corresponding office staff of the target financial project.
A user behavior determination apparatus, comprising:
a request receiving unit, configured to receive a project handling request of a target user, where the project handling request corresponds to a target financial project to be handled, and the project handling request at least includes a handling amount;
a data obtaining unit, configured to obtain item data of the target financial item, where the item data at least includes: the target financial item comprises a preset total amount, a first amount for transacting the target financial item, a second amount for pre-purchasing the target financial item, a first user number for transacting the target financial item and a second user number for pre-purchasing the target financial item;
the model prediction unit is used for inputting the project data into a pre-trained neural network model so as to obtain the predicted residual amount of the target financial project output by the neural network model;
the neural network model is trained by using a plurality of first sample data with residual credit labels, wherein the first sample data comprises a sample total credit of the target financial project, a sample amount of the target financial project, a sample user number of the target financial project and the sample user number of the target financial project;
And the result obtaining unit is used for obtaining a behavior determination result at least according to the predicted remaining sum and the transacting amount, wherein the behavior determination result represents whether the target user transacts the target financial project or not.
An electronic device, comprising:
a memory for storing an application program and data generated by the operation of the application program;
a processor for executing the application program to realize: receiving a project handling request of a target user, wherein the project handling request corresponds to a target financial project to be handled, and the project handling request at least comprises handling amount; obtaining project data of the target financial project, wherein the project data at least comprises: the target financial item comprises a preset total amount, a first amount for transacting the target financial item, a second amount for pre-purchasing the target financial item, a first user number for transacting the target financial item and a second user number for pre-purchasing the target financial item; inputting the project data into a pre-trained neural network model to obtain the predicted residual amount of the target financial project output by the neural network model; the neural network model is trained by using a plurality of first sample data with residual credit labels, wherein the first sample data comprises a sample total credit of the target financial project, a sample amount of the target financial project, a sample user number of the target financial project and the sample user number of the target financial project; and obtaining a behavior determination result at least according to the predicted remaining amount and the transacted amount, wherein the behavior determination result characterizes whether the target user transacts the target financial project.
According to the method, the device and the electronic equipment for determining the user behavior, after receiving the project handling request of the target user, project data of corresponding target financial projects are obtained, wherein the project data comprise a plurality of data items which possibly influence the residual amount of the target financial projects, the project data are input into a pre-trained neural network model, and the neural network model is obtained by training based on historical sample data with residual amount labels, so that the neural network model can predict the residual amount of the target financial projects, and further a behavior determination result representing whether to handle the target financial projects for the target user can be obtained according to the predicted residual amount. Therefore, according to the application, through learning the historical data, prediction is performed through the learned model, so that whether to transact corresponding projects for the user is determined.
Drawings
In order to more clearly illustrate the technical solutions of the embodiments of the present application, the drawings required for the description of the embodiments will be briefly described below, and it is obvious that the drawings in the following description are only some embodiments of the present application, and other drawings may be obtained according to these drawings without inventive effort for a person skilled in the art.
FIG. 1 is a flowchart of a method for determining user behavior according to an embodiment of the present application;
FIG. 2 is a partial flow chart of a method for determining user behavior according to a first embodiment of the present application;
FIG. 3 is another flowchart of a method for determining user behavior according to a first embodiment of the present application;
fig. 4 is a schematic structural diagram of a device for determining user behavior according to a second embodiment of the present application;
fig. 5 and fig. 6 are schematic diagrams of another configuration of a user behavior determining apparatus according to a second embodiment of the present application;
fig. 7 is a schematic structural diagram of an electronic device according to a third embodiment of the present application.
Detailed Description
The following description of the embodiments of the present application will be made clearly and completely with reference to the accompanying drawings, in which it is apparent that the embodiments described are only some embodiments of the present application, but not all embodiments. All other embodiments, which can be made by those skilled in the art based on the embodiments of the application without making any inventive effort, are intended to be within the scope of the application.
Referring to fig. 1, a flowchart of a method for determining user behavior according to an embodiment of the present application is provided, and the method may be applied to an electronic device capable of performing data processing, such as a computer or a server. The technical scheme in the embodiment is mainly used for determining whether to transact corresponding projects for users in the current transacting state of financial projects.
Specifically, the method in this embodiment may include the following steps:
step 101: and receiving a project handling request of the target user.
The project handling request corresponds to a target financial project to be handled, and at least contains handling amount, namely amount of the target financial project which is ready to be purchased by a target user.
Specifically, in this embodiment, the operation performed by the target user on the transaction interface corresponding to the target financial project may be read, and the operation may be analyzed, so as to obtain the project transaction request including the transaction amount that the target user is ready to purchase.
Step 102: project data of a target financial project is obtained.
Wherein, the project data at least comprises: the method comprises the steps of presetting total amount of target financial items, first amount of full money handling target financial items, second amount of pre-purchase handling target financial items, first user number of full money handling target financial items and second user number of pre-purchase handling target financial items.
It should be noted that, the preset total amount of the target financial project refers to: the financial institution can provide the total purchase amount, and the user can purchase the target financial product so as to occupy the amount in the preset total amount; the first amount of the full transaction target financial product is: the total amount of the target financial product is transacted through the full money in the occupied amount in the preset total amount of the target financial product; the second amount of the target financial item of the pre-purchase transaction is different from the fixed amount paid by the target financial item of the pre-purchase transaction, and refers to the amount occupied in the preset total amount of the target financial item, for example, the user A pays an amount of 1000 yuan of the target financial item of the pre-purchase transaction of 100 yuan, and at this time, the second amount refers to 1000 yuan of the preset total amount of the target financial item, instead of the fixed amount of 100 yuan; the first user number of the full money handling target financial product means: the total number of users transacting the target financial product in a full money way; the second user quantity of the pre-purchased transaction target financial product refers to: the total number of users who transact the target financial product in a pre-purchase manner.
Specifically, in this embodiment, the preset total amount, the first amount, the second amount, the first user number and the second user number in the item data may be obtained by performing data collection and data statistics on the current sales state of the target financial item.
Step 103: and inputting the project data into a pre-trained neural network model to obtain the predicted residual amount of the target financial project output by the neural network model.
The neural network model is trained by using a plurality of first sample data with residual credit labels, wherein the first sample data comprises a sample total credit of a target financial item, a sample amount of a full money handling target financial item, a sample amount of a pre-purchase handling target financial item, a sample user number of the full money handling target financial item and a sample user number of the pre-purchase handling target financial item.
It should be noted that, the predicted remaining amount of the target financial project means: the remaining amount of the target financial item available for sale in the event that a user transacting the target financial product with a pre-purchased buyer may violate.
The first sample data in this embodiment may be obtained by collecting and counting data of historical sales states of the target financial product. For example, the sales state data of the target financial item at a plurality of historical moments is collected and counted to obtain a sample total amount of the target financial product at each historical moment, a sample amount of the full money handling target financial item, a sample amount of the pre-purchase transaction target financial item, a sample user number of the full money handling target financial item, and a sample user number of the pre-purchase transaction target financial item, so that first sample data can be obtained, and a historical residual amount of the target financial product at each historical moment can be obtained, and a residual amount label of the first sample data can be obtained.
Based on the above, in this embodiment, a pre-constructed neural network model is trained by using a plurality of first sample data, so as to obtain a trained neural network model, where the neural network model can predict the current residual amount of the target financial project.
Specifically, in this embodiment, after obtaining a plurality of first sample data with residual quota labels, the plurality of first sample data are sequentially input into a neural network model to be trained, model parameters of the neural network model are modified according to predicted residual quota output by the neural network model for each first sample data and corresponding residual quota labels, and modification of the model parameters can be specifically guided by constructing a loss function until the loss function converges, so that training of the neural network model is completed. The specific training mode is referred as follows:
inputting first sample data into a neural network model, obtaining first prediction residual amount of the neural network model, comparing the prediction residual amount with residual amount labels corresponding to the first sample data, obtaining a loss function value of the neural network model according to the residual amount labels corresponding to the prediction residual amount and the first sample data, and modifying model parameters of the neural network model such as parameters in a convolution layer according to the loss function value;
Then, inputting the next first sample data into the neural network model, obtaining a second predicted residual amount of the neural network model, comparing the predicted residual amount with a residual amount label corresponding to the current first sample data, obtaining a loss function value of the neural network model according to the residual amount label corresponding to the predicted residual amount and the current first sample data, and modifying model parameters of the neural network model such as parameters in a convolution layer according to the loss function value, so that the loss function value can be reduced;
and the like, until the loss function value is reduced to be unchanged, the neural network model is stabilized, and the training of the neural network model is completed.
Based on this, in this embodiment, the third sample data may be obtained, and then the neural network model may be tested by using the third sample data, so as to obtain a test result of the neural network model.
Wherein the test results at least characterize the accuracy of the neural network model.
The third sample data is different from the first sample data, and includes: the method comprises the steps of sample total amount of target financial items, sample amount of full money handling the target financial items, sample amount of pre-purchasing handling the target financial items, sample user number of full money handling the target financial items, and sample user number of pre-purchasing handling the target financial items. The third sample data in this embodiment may be obtained by collecting and counting the historical sales status of the target financial product.
Based on the above, the neural network model trained in the embodiment can process the project data and output the corresponding prediction residual result.
Step 104: and obtaining a behavior determination result at least according to the predicted remaining amount and the transaction amount.
Wherein the behavioral determination results characterize whether the target user transacts the target financial project. For example, the behavior determination result is represented by 0 or 1, if the behavior determination result is 0, the target user is not transacted with the target financial item, the total amount of the target financial item is prevented from being excessively increased, and if the behavior determination result is 1, the target user is transacted with the target financial item, so as to meet the financial requirement of the target user.
Specifically, in this embodiment, the predicted remaining amount and the transaction amount may be calculated numerically, so as to obtain a behavior determination result. For example, when the predicted remaining amount is greater than or equal to the transaction amount, a behavior determination result for transacting the target financial item for the target user is generated, and when the predicted remaining amount is less than the transaction amount, a behavior determination result for not transacting the target financial item for the target user is generated.
As can be seen from the above solution, in the method for determining user behavior provided in the first embodiment of the present application, after receiving a request for handling a target user's project, the method includes obtaining project data of a corresponding target financial project, where the project data includes a plurality of data items that may affect a remaining amount of the target financial project, and inputting the data items into a pre-trained neural network model, where the neural network model is obtained by training based on historical sample data with a remaining amount tag, so that the neural network model can predict the remaining amount of the target financial project, and further obtain a behavior determination result that indicates whether to handle the target financial project for the target user according to the predicted remaining amount. Therefore, in this embodiment, by learning the history data, prediction is performed through the learned model, so as to determine whether to transact the corresponding project for the user.
In one implementation, when the behavior determination result is obtained in step 104, the following may be specifically implemented:
and inputting the predicted residual amount, the preset total amount and the transacted amount into a pre-trained classification model to obtain a behavior determination result output by the classification model.
And whether the behavior determination result characterizes the target user transacting the target financial project or not is determined.
Specifically, the classification model is obtained by training a plurality of second sample data with transaction tags, the second sample data comprises a sample total amount, a sample residual amount and a sample transaction amount, and the transaction tags represent whether to transact a target financial item corresponding to the sample transaction amount.
The second sample data in this embodiment may be obtained by performing data collection and statistics on historical sales data of the target financial product. For example, the sales data of the target financial item at a plurality of historical moments is collected and counted to obtain a sample total amount, a sample residual amount and a sample transaction amount of the target financial product at each historical moment, so that second sample data can be obtained, and whether the target financial product is transacted by the user at the sample transaction amount at each historical moment can be obtained, so that a transaction tag of the second sample data can be obtained.
Based on this, in this embodiment, the pre-constructed classification model is trained by using the plurality of second sample data, so as to obtain a trained classification model, and the classification model can determine whether to transact the target financial project for the target user, thereby obtaining the behavior determination result. Based on this, the staff of the financial institution can finally determine whether to transact the target financial product for the target user according to the behavior determination result.
In another implementation manner, when the behavior determination result is obtained in step 104, the method may be specifically implemented as follows, as shown in fig. 2:
step 201: and comparing the processed amount, the predicted remaining amount and the preset total amount to obtain a comparison result.
The comparison result can represent the size or the proportion relation among the transacted amount, the predicted residual amount and the preset total amount.
Step 202: and obtaining a behavior determination result according to the comparison result.
Wherein the behavioral determination results characterize whether the target user transacts the target financial project.
Specifically, when the ratio between the comparison result represents that the processed amount and the preset total amount is smaller than the ratio threshold, if the processed amount is far smaller than the preset total amount, the generated behavior determination result represents that the target user processes the target financial project, and at this time, even if the preset total amount is lifted due to no user default, the situation that the total amount is excessively lifted does not occur.
For example, if the target user needs to purchase the target financial item 100 yuan and the preset total amount is 1 billion, then it may be determined that the target user is transacted with an item transacting request for purchasing the target financial item.
The ratio threshold may be set according to requirements, for example, 10% or 5%, i.e. the preset total amount of the target financial project cannot be excessively increased by 10% or 5% of the amount.
In another case, if the ratio of the comparison result indicates that the processed amount is greater than or equal to the ratio threshold and the processed amount is less than or equal to the predicted remaining amount, the generated behavior determination structure indicates that the target user processes the target financial item, that is, the processed amount is not much less than the preset total amount but the processed amount to be purchased by the target user is less than the predicted remaining amount, at this time, the target financial item has enough amount to provide the target user with the purchase, and thus, it can be determined that the target user processes the item processing request for purchasing the target financial item.
For example, the target user needs to purchase the target financial product 1000 yuan, and the preset total amount is 5 ten thousand, but the predicted remaining amount is 2000 yuan, so that it can be determined that the target user handles the project handling request for purchasing the target financial project.
Further, under the condition that the ratio of the comparison result represents that the processed amount is greater than or equal to a ratio threshold value and the processed amount is greater than the predicted remaining amount and the difference between the processed amount and the predicted remaining amount is less than or equal to a difference threshold value, the generated behavior determination result represents that the target user processes the target financial project;
that is, the transacted amount is not much smaller than the preset total amount and the transacted amount to be purchased by the target user is greater than the predicted remaining amount, but the difference of the transacted amount exceeding the predicted remaining amount is less than or equal to the difference threshold, is not excessively large, and the transacted target transacted item is transacted for the target, but the amount exceeding the preset total amount is not excessively high, and even if there is no user violation, the total amount is not excessively raised, and thus, it can be determined that the target user transacts the item transacted request for purchasing the target transacted item.
And conversely, when the ratio of the comparison result representation processed amount to the preset total amount is greater than or equal to the ratio threshold, the processed amount is greater than the predicted remaining amount, and the difference between the processed amount and the predicted remaining amount is greater than the difference threshold, the generated behavior determination result representation does not process the target financial project for the target user.
That is, the amount of money to be transacted is not much smaller than the preset total amount and the amount of money to be purchased by the target user is larger than the predicted remaining amount, and the difference between the amount of money to be transacted and the predicted remaining amount is excessively large and exceeds the difference threshold.
The difference threshold may be set according to a preset total amount, for example, 5 ten thousand or 10 ten thousand, i.e., the preset total amount of the target financial project cannot be excessively increased by 5 ten thousand or 10 ten thousand.
In one implementation, after step 103, the present embodiment may further include the following steps, as shown in fig. 3:
step 105: and outputting the predicted residual amount of the target financial project to prompt the corresponding office personnel of the target financial project.
The clerk can manually judge whether to transact the target financial project for the target user according to the predicted residual amount prompted.
Specifically, in this embodiment, the predicted remaining amount may be output to the office through a display screen or a mobile phone, etc.
Referring to fig. 4, a schematic structural diagram of a device for determining user behavior according to a second embodiment of the present application may be configured in an electronic device capable of performing data processing, such as a computer or a server. The technical scheme in the embodiment is mainly used for determining whether to transact corresponding projects for users in the current transacting state of financial projects.
Specifically, the apparatus in this embodiment may include the following structures:
a request receiving unit 401, configured to receive a project handling request of a target user, where the project handling request corresponds to a target financial project to be handled, and the project handling request at least includes a handling amount;
a data obtaining unit 402, configured to obtain item data of the target financial item, where the item data at least includes: the target financial item comprises a preset total amount, a first amount for transacting the target financial item, a second amount for pre-purchasing the target financial item, a first user number for transacting the target financial item and a second user number for pre-purchasing the target financial item;
a model prediction unit 403, configured to input the project data into a pre-trained neural network model, so as to obtain a predicted residual amount of the target financial project output by the neural network model;
The neural network model is trained by using a plurality of first sample data with residual credit labels, wherein the first sample data comprises a sample total credit of the target financial project, a sample amount of the target financial project, a sample user number of the target financial project and the sample user number of the target financial project;
and a result obtaining unit 404, configured to obtain a behavior determination result at least according to the predicted remaining sum and the transaction amount, where the behavior determination result characterizes whether to transact the target financial project for the target user.
As can be seen from the above-mentioned scheme, in the determining device for user behavior provided in the second embodiment of the present application, after receiving a request for handling a target user's project, by obtaining project data of a corresponding target financial project, where the project data includes a plurality of data items that may affect the remaining amount of the target financial project, and inputting the project data into a pre-trained neural network model, where the neural network model is obtained by training based on historical sample data with a remaining amount tag, the neural network model can predict the remaining amount of the target financial project, and further obtain a behavior determining result that indicates whether to handle the target financial project for the target user according to the predicted remaining amount. Therefore, in this embodiment, by learning the history data, prediction is performed through the learned model, so as to determine whether to transact the corresponding project for the user.
In one implementation, the result obtaining unit 404 is specifically configured to: inputting the predicted residual amount, the preset total amount and the transacted amount into a pre-trained classification model to obtain a behavior determination result output by the classification model; the behavior determination result characterizes whether the target user transacts the target financial project; the classification model is obtained by training a plurality of second sample data with transaction tags, the second sample data comprises the sample total amount, the sample residual amount and the sample transaction amount, and the transaction tags represent whether to transact a target financial item corresponding to the sample transaction amount.
In one implementation, the result obtaining unit 404 is specifically configured to: comparing the transacted amount, the predicted remaining amount and the preset total amount to obtain a comparison result; and obtaining a behavior determination result according to the comparison result, wherein the behavior determination result characterizes whether the target user handles the target financial project.
And under the condition that the comparison result represents that the ratio of the transacted amount to the preset total amount is smaller than a ratio threshold, the behavior determination result represents that the target user transacts the target financial project.
And the behavior determination result is characterized in that the target user handles the target financial project under the condition that the comparison result represents that the ratio of the handled amount to the preset total amount is larger than or equal to the ratio threshold and the handled amount is smaller than or equal to the predicted remaining amount.
Further, the behavior determination result is characterized in that the target user handles the target financial project when the comparison result indicates that the ratio of the handled amount to the preset total amount is greater than or equal to the ratio threshold, the handled amount is greater than the predicted remaining amount, and the difference between the handled amount and the predicted remaining amount is less than or equal to the difference threshold;
and under the condition that the comparison result represents that the ratio of the transacted amount to the preset total amount is larger than or equal to the ratio threshold, the transacted amount is larger than the predicted remaining amount, and the difference between the transacted amount and the predicted remaining amount is larger than the difference threshold, the behavior determination result represents that the target financial project is not transacted for the target user.
In one implementation, the apparatus in this embodiment may further include the following structure, as shown in fig. 5:
A model test unit 405, configured to: obtaining third sample data, the third sample data comprising: the sample total amount of the target financial item, the sample user number of the target financial item, and the sample user number of the target financial item; testing the neural network model by using the third sample data to obtain a test result of the neural network model; the test results at least characterize the accuracy of the neural network model.
In one implementation, the apparatus in this embodiment may further include the following structure, as shown in fig. 6:
a prediction output unit 406 for: and outputting the predicted residual amount of the target financial project to prompt the corresponding office staff of the target financial project.
It should be noted that, the specific implementation of each unit in this embodiment may refer to the corresponding content in the foregoing, which is not described in detail herein.
Referring to fig. 7, a schematic structural diagram of an electronic device according to a third embodiment of the present application may be suitable for an electronic device capable of performing data processing, such as a computer or a server. The technical scheme in the embodiment is mainly used for determining whether to transact corresponding projects for users in the current transacting state of financial projects.
Specifically, the electronic device in this embodiment may include the following structure:
a memory 701 for storing an application program and data generated by the application program running;
a processor 702 for executing an application to implement: receiving a project handling request of a target user, wherein the project handling request corresponds to a target financial project to be handled, and the project handling request at least comprises handling amount; obtaining project data of the target financial project, wherein the project data at least comprises: the target financial item comprises a preset total amount, a first amount for transacting the target financial item, a second amount for pre-purchasing the target financial item, a first user number for transacting the target financial item and a second user number for pre-purchasing the target financial item; inputting the project data into a pre-trained neural network model to obtain the predicted residual amount of the target financial project output by the neural network model; the neural network model is trained by using a plurality of first sample data with residual credit labels, wherein the first sample data comprises a sample total credit of the target financial project, a sample amount of the target financial project, a sample user number of the target financial project and the sample user number of the target financial project; and obtaining a behavior determination result at least according to the predicted remaining amount and the transacted amount, wherein the behavior determination result characterizes whether the target user transacts the target financial project.
As can be seen from the above-mentioned scheme, in the electronic device provided in the third embodiment of the present application, after receiving a request for handling a target user's project, the electronic device obtains project data of a corresponding target financial project, where the project data includes a plurality of data items that may affect the remaining amount of the target financial project, and inputs the project data into a pre-trained neural network model, where the neural network model is obtained by training based on historical sample data with a remaining amount tag, so that the neural network model can predict the remaining amount of the target financial project, and further obtain a behavior determination result that indicates whether to handle the target financial project for the target user according to the predicted remaining amount. Therefore, in this embodiment, by learning the history data, prediction is performed through the learned model, so as to determine whether to transact the corresponding project for the user.
It should be noted that, the specific implementation of the processor in this embodiment may refer to the corresponding content in the foregoing, which is not described in detail herein.
Based on the above implementation scheme, taking a scenario that a banking staff handles a target financial product for a user as an example, the technical scheme of the application is illustrated:
Firstly, banks are difficult to store, in order to improve the storage capacity, users who pay money on hands but cannot take enough funds for buying certain funds for a fixed length of time develop a pre-purchase function, namely, customers can purchase financial products (some financial products have the limitation of purchase amount) first, only need to pay other amounts at fixed time, so that the storage capacity of the banks can be improved, but some customers always violate the limit occupied before the money is released, so that the financial products cannot be used by people, and the amount of the banks is wasted.
In order to solve the problems, in a specific implementation, a bank financial product line recommendation system is established based on the technical scheme of the application, the total line of the financial product is reasonably improved by analyzing the default condition of the past financial product of the bank, the total line of the financial product can be ensured to be basically ensured under the default condition of a client, the expected line of the bank can be ensured, and the user experience is improved while the bank benefit is ensured. The method comprises the following steps:
the system is realized by collecting total line of the pre-purchased financial product, total purchase amount of the pre-purchased money, total amount of the pre-purchased customers, total number of the pre-purchased customers, total amount of the total purchased money and total amount left after the customer violations as historical data to train and verify the neural network model, taking the total line of the pre-purchased financial product, total purchase amount of the pre-purchased customers, total amount of the total purchased money as output of the neural network model, combining BP (back propagation) neural network and genetic algorithm GA (Genetic Algorithm) in the neural network model, and introducing genetic algorithm in the optimization aspect of the weight and threshold value of the BP neural network to construct the GA-BP neural network model. In the GA-BP neural network model, the structure of the BP neural network is determined according to the number of network inputs and outputs, and then the number of parameters to be optimized in a genetic algorithm is determined.
The GA-BP neural network model can be a three-layer model and comprises an input layer, a hidden layer and an output layer, wherein the number of hidden layer nodes is determined by adopting a trial-and-error method, so that the structure of the GA-BP neural network is determined. And taking the optimal individual output by the genetic algorithm as an initial weight and a threshold value of the BP neural network to train and learn the BP neural network.
Specifically, based on the analysis of the above historical data, the GA-BP neural network model is trained, and the prediction accuracy of the model is verified by using a test sample, so that the available neural network model is obtained.
Based on the above, when a new user purchases a financial product, a bank staff inputs whether the user is full money purchase or pre-purchase and purchase amount in the system realized by the application, the system realized by the application inquires and calculates the total amount of the financial product, the total amount of full money purchase amount (if the current user is full money purchase, the system is automatically added), the total amount of pre-purchased user (if the current user is pre-purchase, the system is automatically added), the number of pre-purchased user and the number of full money purchased user into a neural network model, and can output two output to the bank staff, wherein one is the total amount of the financial product, the other is the total amount remained after the customer predicted by the model breaks, and the bank staff judges whether to transact the service for the customer according to the output result.
Furthermore, the system realized by the application inputs the predicted residual total amount, the predicted total amount of the financial product and the predicted purchase amount into a pre-trained classification model, and the classification model can process the input data through the pre-trained history data learning so as to further estimate whether to transact the business of purchasing the financial product for a new user.
Therefore, according to the historical data, the application predicts the residual amount and the amount of the amount improvement of the financial product by means of the neural network model, and ensures that the total amount of the bank can still be kept at the expected level of the bank under the condition that the customer breaks the contract.
In the present specification, each embodiment is described in a progressive manner, and each embodiment is mainly described in a different point from other embodiments, and identical and similar parts between the embodiments are all enough to refer to each other. For the device disclosed in the embodiment, since it corresponds to the method disclosed in the embodiment, the description is relatively simple, and the relevant points refer to the description of the method section.
Those of skill would further appreciate that the various illustrative elements and algorithm steps described in connection with the embodiments disclosed herein may be implemented as electronic hardware, computer software, or combinations of both, and that the various illustrative elements and steps are described above generally in terms of functionality in order to clearly illustrate the interchangeability of hardware and software. Whether such functionality is implemented as hardware or software depends upon the particular application and design constraints imposed on the solution. Skilled artisans may implement the described functionality in varying ways for each particular application, but such implementation decisions should not be interpreted as causing a departure from the scope of the present application.
The steps of a method or algorithm described in connection with the embodiments disclosed herein may be embodied directly in hardware, in a software module executed by a processor, or in a combination of the two. The software modules may be disposed in Random Access Memory (RAM), memory, read Only Memory (ROM), electrically programmable ROM, electrically erasable programmable ROM, registers, hard disk, a removable disk, a CD-ROM, or any other form of storage medium known in the art.
The previous description of the disclosed embodiments is provided to enable any person skilled in the art to make or use the present application. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the generic principles defined herein may be applied to other embodiments without departing from the spirit or scope of the application. Thus, the present application is not intended to be limited to the embodiments shown herein but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.
Claims (10)
1. A method for determining user behavior, comprising:
receiving a project handling request of a target user, wherein the project handling request corresponds to a target financial project to be handled, and the project handling request at least comprises handling amount;
Obtaining project data of the target financial project, wherein the project data at least comprises: the target financial item comprises a preset total amount, a first amount for transacting the target financial item, a second amount for pre-purchasing the target financial item, a first user number for transacting the target financial item and a second user number for pre-purchasing the target financial item;
inputting the project data into a pre-trained neural network model to obtain the predicted residual amount of the target financial project output by the neural network model; the predicted remaining amount of the target financial project refers to: a remaining amount of the target financial item available for sale in the event that a user transacting the target financial item with a pre-purchased buyer may violate the target financial item;
the neural network model is trained by using a plurality of first sample data with residual credit labels, wherein the first sample data comprises a sample total credit of the target financial project, a sample amount of the target financial project, a sample user number of the target financial project and the sample user number of the target financial project;
And obtaining a behavior determination result at least according to the predicted remaining amount and the transacted amount, wherein the behavior determination result characterizes whether the target user transacts the target financial project.
2. The method of claim 1, wherein obtaining a behavioral determination based at least on the predicted remaining credit and the transaction amount comprises:
inputting the predicted residual amount, the preset total amount and the transacted amount into a pre-trained classification model to obtain a behavior determination result output by the classification model; the behavior determination result characterizes whether the target user transacts the target financial project;
the classification model is obtained by training a plurality of second sample data with transaction tags, the second sample data comprises the sample total amount, the sample residual amount and the sample transaction amount, and the transaction tags represent whether to transact a target financial item corresponding to the sample transaction amount.
3. The method of claim 1, wherein obtaining a behavioral determination based at least on the predicted remaining credit and the transaction amount comprises:
Comparing the transacted amount, the predicted remaining amount and the preset total amount to obtain a comparison result;
and obtaining a behavior determination result according to the comparison result, wherein the behavior determination result characterizes whether the target user handles the target financial project.
4. The method of claim 3, wherein the behavioral determination is characterized as the target user transacting the target financial item if the comparison characterizes a ratio of the transacted amount to the preset total amount less than a ratio threshold.
5. The method of claim 3, wherein the behavioral determination is characterized as the target user transacting the target financial item if the comparison characterizes a ratio of the transacted amount to the preset total amount greater than or equal to a ratio threshold and the transacted amount less than or equal to the predicted remaining amount.
6. The method according to claim 3, wherein the behavior determination result is characterized in that the target user deals with the target financial item in a case where the comparison result characterizes that a ratio of the amount of business to the preset total amount is greater than or equal to a ratio threshold, the amount of business is greater than the predicted remaining amount, and a difference between the amount of business and the predicted remaining amount is less than or equal to a difference threshold;
And under the condition that the comparison result represents that the ratio of the transacted amount to the preset total amount is larger than or equal to the ratio threshold, the transacted amount is larger than the predicted remaining amount, and the difference between the transacted amount and the predicted remaining amount is larger than the difference threshold, the behavior determination result represents that the target financial project is not transacted for the target user.
7. The method as recited in claim 1, further comprising:
obtaining third sample data, the third sample data comprising: the sample total amount of the target financial item, the sample user number of the target financial item, and the sample user number of the target financial item;
testing the neural network model by using the third sample data to obtain a test result of the neural network model; the test results at least characterize the accuracy of the neural network model.
8. The method as recited in claim 1, further comprising:
and outputting the predicted residual amount of the target financial project to prompt the corresponding office staff of the target financial project.
9. A user behavior determining apparatus, comprising:
a request receiving unit, configured to receive a project handling request of a target user, where the project handling request corresponds to a target financial project to be handled, and the project handling request at least includes a handling amount;
a data obtaining unit, configured to obtain item data of the target financial item, where the item data at least includes: the target financial item comprises a preset total amount, a first amount for transacting the target financial item, a second amount for pre-purchasing the target financial item, a first user number for transacting the target financial item and a second user number for pre-purchasing the target financial item;
the model prediction unit is used for inputting the project data into a pre-trained neural network model so as to obtain the predicted residual amount of the target financial project output by the neural network model; the predicted remaining amount of the target financial project refers to: a remaining amount of the target financial item available for sale in the event that a user transacting the target financial item with a pre-purchased buyer may violate the target financial item;
The neural network model is trained by using a plurality of first sample data with residual credit labels, wherein the first sample data comprises a sample total credit of the target financial project, a sample amount of the target financial project, a sample user number of the target financial project and the sample user number of the target financial project;
and the result obtaining unit is used for obtaining a behavior determination result at least according to the predicted remaining sum and the transacting amount, wherein the behavior determination result represents whether the target user transacts the target financial project or not.
10. An electronic device, comprising:
a memory for storing an application program and data generated by the operation of the application program;
a processor for executing the application program to realize: receiving a project handling request of a target user, wherein the project handling request corresponds to a target financial project to be handled, and the project handling request at least comprises handling amount; obtaining project data of the target financial project, wherein the project data at least comprises: the target financial item comprises a preset total amount, a first amount for transacting the target financial item, a second amount for pre-purchasing the target financial item, a first user number for transacting the target financial item and a second user number for pre-purchasing the target financial item; inputting the project data into a pre-trained neural network model to obtain the predicted residual amount of the target financial project output by the neural network model; the predicted remaining amount of the target financial project refers to: a remaining amount of the target financial item available for sale in the event that a user transacting the target financial item with a pre-purchased buyer may violate the target financial item; the neural network model is trained by using a plurality of first sample data with residual credit labels, wherein the first sample data comprises a sample total credit of the target financial project, a sample amount of the target financial project, a sample user number of the target financial project and the sample user number of the target financial project; and obtaining a behavior determination result at least according to the predicted remaining amount and the transacted amount, wherein the behavior determination result characterizes whether the target user transacts the target financial project.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202011048577.1A CN112116169B (en) | 2020-09-29 | 2020-09-29 | User behavior determining method and device and electronic equipment |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202011048577.1A CN112116169B (en) | 2020-09-29 | 2020-09-29 | User behavior determining method and device and electronic equipment |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN112116169A CN112116169A (en) | 2020-12-22 |
| CN112116169B true CN112116169B (en) | 2023-09-19 |
Family
ID=73797829
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202011048577.1A Active CN112116169B (en) | 2020-09-29 | 2020-09-29 | User behavior determining method and device and electronic equipment |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN112116169B (en) |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109117980A (en) * | 2018-06-29 | 2019-01-01 | 阿里巴巴集团控股有限公司 | Data pre-fill method, device and equipment |
| CN109255506A (en) * | 2018-11-22 | 2019-01-22 | 重庆邮电大学 | A kind of internet finance user's overdue loan prediction technique based on big data |
| CN111192131A (en) * | 2019-12-12 | 2020-05-22 | 上海淇玥信息技术有限公司 | Financial risk prediction method and device and electronic equipment |
| CN111667307A (en) * | 2020-05-28 | 2020-09-15 | 中国工商银行股份有限公司 | Method and device for predicting financial product sales volume |
Family Cites Families (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20190259095A1 (en) * | 2018-02-20 | 2019-08-22 | Southwest Business Corporation | Determining present and future virtual balances for a client computing device |
-
2020
- 2020-09-29 CN CN202011048577.1A patent/CN112116169B/en active Active
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109117980A (en) * | 2018-06-29 | 2019-01-01 | 阿里巴巴集团控股有限公司 | Data pre-fill method, device and equipment |
| CN109255506A (en) * | 2018-11-22 | 2019-01-22 | 重庆邮电大学 | A kind of internet finance user's overdue loan prediction technique based on big data |
| CN111192131A (en) * | 2019-12-12 | 2020-05-22 | 上海淇玥信息技术有限公司 | Financial risk prediction method and device and electronic equipment |
| CN111667307A (en) * | 2020-05-28 | 2020-09-15 | 中国工商银行股份有限公司 | Method and device for predicting financial product sales volume |
Also Published As
| Publication number | Publication date |
|---|---|
| CN112116169A (en) | 2020-12-22 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US8666847B1 (en) | Methods systems and computer program products for monitoring inventory and prices | |
| KR101794221B1 (en) | System and method for providing calculation of online sellers | |
| CN106097044A (en) | A kind of data recommendation processing method and device | |
| CN107103548A (en) | The monitoring method and system and risk monitoring and control method and system of network behavior data | |
| WO2020111070A1 (en) | Computer system, method, and program for remitting funds to retailer during electronic commercial transaction | |
| CN113919886A (en) | Data characteristic combination pricing method and system based on summer pril value and electronic equipment | |
| CN108182587A (en) | A kind of electric business platform brush single act detection method and system | |
| KR101089183B1 (en) | Service Method for Loan using of Trade Receivables Collateral by Online Shopping mall | |
| CN113052705A (en) | Supply chain financial service process management method and device | |
| CN110458691A (en) | Risk monitoring and control method and device before a kind of loan | |
| US20210158439A1 (en) | Method for Optimizing Usage Savings in Financial Accounts | |
| CN113269629A (en) | Credit limit determining method, electronic equipment and related product | |
| CN112116169B (en) | User behavior determining method and device and electronic equipment | |
| CN111144899A (en) | Method and device for identifying false transactions and electronic equipment | |
| CN116645134A (en) | Method, device, equipment and medium for recommending credit card in stages | |
| CN112381617B (en) | Online shopping method and device | |
| CN114004660A (en) | Data processing method and device, electronic equipment and storage medium | |
| CN113393328A (en) | Method and device for assessing pre-financing and pre-loan approval and computer storage medium | |
| CN110084370A (en) | A kind of user behavior analysis method and its server | |
| US12014430B2 (en) | Time-based input and output monitoring and analysis to predict future inputs and outputs | |
| JP6836255B2 (en) | Computer systems, methods, and programs for funding retailers in e-commerce | |
| CN117873906B (en) | Method and device for testing prize amount distribution in transaction system | |
| US11004146B1 (en) | Business health score and prediction of credit worthiness using credit worthiness of customers and vendors | |
| US20230351294A1 (en) | Forecasting periodic inter-entity exposure for prophylactic mitigation | |
| US20230401417A1 (en) | Leveraging multiple disparate machine learning model data outputs to generate recommendations for the next best action |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |