CN111951008A - A risk prediction method, apparatus, electronic device and readable storage medium - Google Patents

Abstract

The present application relates to the field of electronic devices, and in particular, to a method, an apparatus, an electronic device, and a readable storage medium for risk prediction, where the method includes: by collecting the transaction information of the target user; determining risk influence factor information according to the target user transaction information; inputting the risk influence factor information into a risk prediction model; the risk prediction model is a convolutional neural network model loaded with a group intelligent optimization algorithm; and outputting a risk prediction result by a risk prediction model, wherein the risk prediction result is used for representing whether the target user is a risk user. The risk prediction scheme based on the CNN model loaded with the BSA algorithm disclosed by the embodiment of the application can be used for rapidly, efficiently and accurately predicting whether a user is a risk user.

Description

A risk prediction method, apparatus, electronic device and readable storage medium

technical field

The present application relates to the technical field of information security, and in particular, to a risk prediction method, apparatus, electronic device and readable storage medium.

Background technique

In order to attract new users and cultivate users' consumption habits in the fierce competition, various types of marketing activities and subsidy activities are emerging one after another. While bringing benefits to normal users, they also spawned a group of black products that focus on marketing activities. Users, the so-called "wool party". At present, under the temptation of huge interests, the methods and technologies of the woolen party are getting faster and faster, and the traditional risk control system based on expert rules has been difficult to keep up with the iteration of the woolen method. Only after making a profit can targeted online rules be used for risk control detection. In this way, it is easy to form a vicious circle of "profit from scouring wool - deployment rules - profit from scouring wool changing methods twice - adjusting rules", and it is impossible to fundamentally detect and prevent the risk of scouring wool. In the prior art, there are also schemes of using statistical methods or neural networks to discover malicious users or IPs, but there are disadvantages such as low efficiency, high false positive rate, and high false negative rate in identifying network black production data.

SUMMARY OF THE INVENTION

The purpose of this application is to solve at least one of the above-mentioned technical defects. The technical scheme adopted in this application is as follows:

In a first aspect, an embodiment of the present application provides a risk prediction method, the method includes:

Collect transaction information of target users;

Determine risk influencing factor information according to the target user transaction information;

Inputting the risk influencing factor information into a risk prediction model; wherein the risk prediction model is a convolutional neural network model loaded with a swarm intelligence optimization algorithm;

The risk prediction model outputs a risk prediction result, wherein the risk prediction result is used to characterize whether the target user is a risk user.

Optionally, the swarm intelligence optimization algorithm is a bird flock optimization algorithm.

Optionally, the construction of the risk prediction model includes:

Loading the bird flock optimization algorithm into the convolutional neural network model, and determining the parameters of the convolutional neural network model according to the bird flock optimization algorithm;

Obtain sample transaction information to train the neural network model loaded with the bird flock optimization algorithm;

According to the preset training rules, the trained neural network model loaded with the bird flock optimization algorithm is determined as the risk prediction model.

Optionally, the bird flock optimization algorithm is a bird flock optimization algorithm with boundary constraints added.

Optionally, the preset rules include at least one of the following:

When the number of training iterations reaches a preset threshold, it is determined that the convolutional neural network model loaded with the bird flock optimization algorithm is a risk prediction model;

or,

Input the sample transaction data to the convolutional neural network model loaded with the bird flock optimization algorithm and iteratively train a predetermined number of times to calculate the sample fitness value;

If the sample fitness value conforms to the preset fitness value, the training is stopped, and the convolutional neural network model loaded with the bird flock optimization algorithm is determined as a risk prediction model.

Optionally, the method includes: collecting target user transaction data in real time.

Second aspect An embodiment of the present invention provides a risk prediction device, the device includes: a collection module, a determination module, an input module, a prediction module, and a storage module, wherein:

The collection module is used to collect target user transaction information;

The determining module is configured to determine risk influencing factor information according to the target user transaction information;

the input module, configured to input the risk influencing factor information into a risk prediction model;

The storage module is used to store a risk prediction model, wherein the risk prediction model is a convolutional neural network model loaded with a swarm intelligence optimization algorithm;

The prediction module is configured to control the risk prediction model to output a risk prediction result, wherein the risk prediction result is used to characterize whether the target user is a risk user.

Optionally, the swarm intelligence optimization algorithm is a bird swarm optimization algorithm with boundary constraints added.

Optionally, the apparatus further includes a model building module, wherein the model building module is used for:

Loading the bird flock optimization algorithm into the convolutional neural network model, and determining the parameters of the convolutional neural network model according to the bird flock optimization algorithm;

Obtain sample transaction information to train the neural network model loaded with the bird flock optimization algorithm;

According to the preset training rules, the trained neural network model loaded with the bird flock optimization algorithm is determined as the risk prediction model.

Optionally, the collection module is further configured to collect transaction data of target users in real time.

In a third aspect, an embodiment of the present invention provides an electronic device, including a processor and a memory;

the memory for storing operation instructions;

The processor is configured to execute the above risk prediction method by invoking the operation instruction.

In a fourth aspect, a computer-readable storage medium is provided with a computer program stored thereon, and when the computer program is executed by a processor, the above-mentioned risk prediction method is implemented.

The beneficial effects brought by the technical solutions provided in the embodiments of the present application are: collecting target user transaction information; determining risk influencing factor information according to the target user transaction information; inputting the risk influencing factor information into a risk prediction model; wherein the The risk prediction model is a convolutional neural network model loaded with a swarm intelligence optimization algorithm; the risk prediction model outputs a risk prediction result, wherein the risk prediction result is used to characterize whether the target user is a risk user. The risk prediction scheme disclosed in the embodiment of the present application uses the convolutional neural network model of the improved bird flock optimization algorithm to process and predict user transaction data, which can quickly, efficiently and accurately predict whether a user is a risk user, and has the advantages of high accuracy and low false alarm rate. , the characteristics of low cost. Avoid risk users, such as black-produced users, who use marketing plan loopholes to arbitrage to bring unnecessary losses to merchants or customers,

Description of drawings

In order to illustrate the technical solutions in the embodiments of the present application more clearly, the following briefly introduces the accompanying drawings that need to be used in the description of the embodiments of the present application.

1 is a schematic flowchart of a risk prediction method provided by an embodiment of the present application;

FIG. 2 is a schematic structural diagram of a risk prediction device provided by an embodiment of the present application;

FIG. 3 is a schematic structural diagram of an electronic device according to an embodiment of the present application.

Detailed ways

The following describes in detail the embodiments of the present application, examples of which are illustrated in the accompanying drawings, wherein the same or similar reference numerals refer to the same or similar elements or elements having the same or similar functions throughout. The embodiments described below with reference to the accompanying drawings are exemplary and are only used to explain the present application, but not to be construed as limiting the present invention.

It will be understood by those skilled in the art that the singular forms "a", "an", "the" and "the" as used herein can include the plural forms as well, unless expressly stated otherwise. It should be further understood that the word "comprising" used in the specification of this application refers to the presence of stated features, integers, steps, operations, elements and/or components, but does not preclude the presence or addition of one or more other features, Integers, steps, operations, elements, components and/or groups thereof. As used herein, the term "and/or" includes all or any element and all combination of one or more of the associated listed items.

The risk prediction technology involved in the embodiments of the present application, specifically relates to a scheme of risk prediction for a convolutional neural network model loaded with a bird flock optimization algorithm, which can be used to detect illegal users. In order to introduce the embodiments of the present application more clearly, the following To introduce some definitions, concepts or devices that may be used in the embodiments:

Black-produced data: data such as account numbers, devices, mobile phone numbers, and locations used by fraudulent gangs.

Black trade users: Black trade users who take the Internet as the medium and network technology as the main means to bring major threats to the security of computer information systems and the order of cyberspace management.

False transaction: The buyer and the seller do not have the actual purchase of goods.

Pick up wool: Take advantage of various promotions to gain benefits.

Bird swarm optimization algorithm, English for Bird Swarm Algorithom, referred to as BSA algorithm, is a swarm intelligence algorithm.

Convolutional Neural Networks (CNN, hereinafter referred to as CNN model) is a kind of feedforward neural network (Feedforward Neural Networks) that includes convolution calculation and has a deep structure. It is one of the representative algorithms of deep learning. . Convolutional neural network has the ability of representation learning and can perform shift-invariant classification of input information according to its hierarchical structure.

The technical solution of the present application and how the technical solution of the present application solves the above-mentioned technical problems will be described in detail below with specific embodiments in conjunction with the accompanying drawings. The following specific embodiments may be combined with each other, and the same or similar concepts or processes may not be repeated in some embodiments. The embodiments of the present application will be described below with reference to the accompanying drawings.

In order to make the purpose, technical solutions and advantages of the present application clearer, FIG. 1 discloses a flowchart of a risk prediction method provided by an embodiment of the present application. As shown in FIG. 1 , the risk prediction method includes:

S101. Collect target user transaction information;

S102, determining risk influencing factor information according to the target user transaction information;

S103, inputting the risk influencing factor information into a risk prediction model; wherein the risk prediction model is a convolutional neural network model loaded with a swarm intelligence optimization algorithm;

S104. The risk prediction model outputs a risk prediction result, wherein the risk prediction result is used to characterize whether the target user is a risk user.

In an optional embodiment, the swarm intelligent optimization algorithm may be a bird swarm optimization algorithm, a cluster optimization algorithm, a particle swarm optimization algorithm and other intelligent optimization algorithms.

The risk prediction method disclosed in this embodiment is mainly implemented based on a risk prediction model. Therefore, the risk prediction model and the construction of the risk prediction model are preliminarily introduced in this embodiment.

In the embodiment of the present application, the risk prediction mode is a convolutional neural network model loaded with a swarm intelligence optimization algorithm. Optionally, when the intelligent swarm optimization algorithm is a bird swarm optimization algorithm, the solution for constructing a risk prediction model includes:

Step 1. Load the bird flock optimization algorithm into the convolutional neural network model, and determine the parameters of the convolutional neural network model according to the bird flock optimization algorithm, that is, determine the hidden layer parameters of the CNN model.

Step 2: Obtain sample transaction information to train the neural network model loaded with the bird flock optimization algorithm; the sample transaction information is transaction information of multiple users including risky users, and the risk influencing factors are extracted from the sample transaction information Information (also known as eigenvalue information), calculate the influence coefficient of the risk influencing factor, and input the risk influencing factor into the CNN model loaded with the BSA algorithm for training.

Step 3. Determine the trained neural network model loaded with the bird flock optimization algorithm as the risk prediction model according to the preset training rules:

In this step, the preset training rule can be set to use the sample transaction data to train the CNN model loaded with the BSA algorithm when the number of iterations reaches a preset threshold. For example, if the training iteration is 500, it is considered that the model has reached a stable state. It can be determined that the convolutional neural network model loaded with the bird flock optimization algorithm is a risk prediction model;

Optionally, the preset training rules can also be set to the following rules:

Step 3-1. Input the sample transaction data to the convolutional neural network model loaded with the bird flock optimization algorithm, and calculate the sample fitness value after iterative training for a predetermined number of times: divide the sample transaction data into 10 parts according to the ten-fold cross method, of which 9 The CNN model loaded with the BSA algorithm is trained by using the above 10 training data respectively, and the results obtained in the first 9 copies are used as the training results and the verification results of the verification data are compared. to adjust the model. The training process is as follows: input the training data into the CNN model loaded with the BSA algorithm, and iterate a predetermined number of times. For example, after one iteration, update the position information of the individual birds, adjust the parameters of the hidden layer of the model, and then start the iteration. Train and calculate the fitness value of individual birds, and update the current optimal fitness value and optimal performance.

Step 3-2, if the sample fitness value meets the preset fitness value, stop training, and determine that the convolutional neural network model loaded with the bird flock optimization algorithm is a risk prediction model: when calculating in step 3-2 The optimal fitness value of the model conforms to the preset fitness value. For example, if the accuracy rate of the verification result of the model reaches 97%, and the recall rate of the model, that is, the recall rate reaches more than 95%, the model is considered to have reached a stable state.

Optionally, when the sample fitness value conforms to the preset fitness value, it may be determined whether the current number of iterations has reached the predetermined number of iterations, and if not, iterative training is performed again.

In an optional embodiment of the present application, the bird flock optimization algorithm is a bird flock optimization algorithm with boundary constraints added, thereby improving the quality of the solution and the algorithm convergence speed and avoiding falling into a local optimum.

The basic implementation process of risk prediction based on the risk prediction model constructed in the above embodiment will be described below:

Step 1, collect target user transaction messages in real time;

Step 2. Determine the risk influencing factor information according to the target user transaction information: classify the transaction messages collected in real time in step 1, extract the risk influencing factor and calculate the influence coefficient, and the risk influencing factor includes but is not limited to: the target user The number of logins in the predetermined period, the number of registered accounts of the target user (or target terminal), the number of times the transaction is initiated to the same object within the predetermined period, the number of times the transaction amount exceeds the predetermined upper limit in the predetermined period, the target user's recent transaction location information, the target user There are 218 risk factors, such as the number of transactions on a single trading day, the total number of transactions in the most recent predetermined period, the level of transaction objects and historical objects, location information and whether the mobile banking contracted merchants are located in different places.

Step 3. Input all the obtained risk influencing factor information into the risk prediction model;

Step 4: The risk prediction model outputs a risk prediction result, wherein the risk prediction result is used to characterize whether the target user is a risk user. For example, when the prediction result output by the risk prediction model is 1, the target user is considered to be a risk user, such as a wool party, otherwise it is a non-risk user.

Step 5: Send the risk prediction result of the target user to a third-party platform and other user analysis platforms that have been deployed for downstream applications.

The risk prediction scheme based on the CNN model loaded with the BSA algorithm disclosed in the embodiment of the present application can quickly, efficiently and accurately predict whether a user is a risk user, and can effectively trace the source of the risk user and process it, with high accuracy and false alarm rate. low cost and low cost.

FIG. 2 shows that an embodiment of the present application provides a risk prediction device. As shown in FIG. 2 , the device may mainly include: 201 a collection module, 202 a determination module, 203 an input module, 204 a storage module, 205 a prediction module and, wherein :

The 201 collection module is used to collect target user transaction information;

The 202 determination module is configured to determine risk impact factor information according to the target user transaction information;

The 203 input module is used to input the risk impact factor information into a risk prediction model;

The 204 storage module is used to store a risk prediction model, wherein the risk prediction model is a convolutional neural network model loaded with a swarm intelligence optimization algorithm;

The prediction module 205 is configured to control the risk prediction model to output a risk prediction result, wherein the risk prediction result is used to characterize whether the target user is a risk user.

In an optional embodiment, the swarm intelligence optimization algorithm is a bird swarm optimization algorithm with boundary constraints added.

In an optional embodiment, the apparatus further includes a model building module, wherein the model building module is used for:

Loading the bird flock optimization algorithm into the convolutional neural network model, and determining the parameters of the convolutional neural network model according to the bird flock optimization algorithm;

Obtain sample transaction information to train the neural network model loaded with the bird flock optimization algorithm;

According to the preset training rules, the trained neural network model loaded with the bird flock optimization algorithm is determined as the risk prediction model.

In an optional embodiment, the collection module is also used to collect target user transaction data in real time.

It can be understood that, the above-mentioned modules of the risk prediction apparatus in this embodiment have the function of implementing the corresponding steps of the method in the embodiment shown in FIG. 1 . This function can be implemented by hardware or by executing corresponding software by hardware. The hardware or software includes one or more modules corresponding to the above functions. The above-mentioned modules may be software and/or hardware, and the above-mentioned modules may be implemented independently, or multiple modules may be integrated and implemented. For the functional description of the above modules, reference may be made to the corresponding description of the method in the embodiment shown in FIG. 1 , and details are not repeated here.

Embodiments of the present application provide an electronic device, including a processor and a memory;

memory for storing operation instructions;

The processor is configured to execute the risk prediction method provided in any embodiment of the present application by invoking the operation instruction.

As an example, FIG. 3 shows a schematic structural diagram of an electronic device to which an embodiment of the present application is applied. As shown in FIG. 3 , the electronic device 2000 includes: a processor 2001 and a memory 2003 . The processor 2001 is connected to the memory 2003, for example, through the bus 2002. Optionally, the electronic device 2000 may further include a transceiver 2004 . It should be noted that, in practical applications, the transceiver 2004 is not limited to one, and the structure of the electronic device 2000 does not constitute a limitation to the embodiments of the present application.

The processor 2001 is used in the embodiments of the present application to implement the methods shown in the foregoing method embodiments. The transceiver 2004 may include a receiver and a transmitter, and the transceiver 2004 is applied in the embodiments of the present application, and is configured to implement the function of communicating between the electronic device and other devices in the embodiments of the present application during execution.

The processor 2001 may be a CPU (Central Processing Unit, central processing unit), a general-purpose processor, a DSP (Digital Signal Processor, data signal processor), an ASIC (Application Specific Integrated Circuit, an application-specific integrated circuit), an FPGA (Field Programmable Gate Array, Field Programmable Gate Array) or other programmable logic devices, transistor logic devices, hardware components, or any combination thereof. It may implement or execute the various exemplary logical blocks, modules and circuits described in connection with this disclosure. The processor 2001 may also be a combination that implements computing functions, such as a combination of one or more microprocessors, a combination of a DSP and a microprocessor, and the like.

The bus 2002 may include a path to communicate information between the components described above. The bus 2002 may be a PCI (Peripheral Component Interconnect, Peripheral Component Interconnect Standard) bus or an EISA (Extended Industry Standard Architecture, Extended Industry Standard Architecture) bus or the like. The bus 2002 can be divided into an address bus, a data bus, a control bus, and the like. For ease of presentation, only one thick line is used in FIG. 3, but it does not mean that there is only one bus or one type of bus.

The memory 2003 may be ROM (Read Only Memory, read only memory) or other types of static storage devices that can store static information and instructions, RAM (Random Access Memory, random access memory) or other types of storage information and instructions. A dynamic storage device can also be an EEPROM (Electrically Erasable Programmable Read Only Memory), a CD-ROM (Compact Disc Read Only Memory, a CD-ROM) or other CD-ROM storage, CD-ROM storage (including compressed CDs, Laser Disc, Optical Disc, Digital Versatile Disc, Blu-ray Disc, etc.), magnetic disk storage medium or other magnetic storage device, or any other capable of carrying or storing desired program code in the form of instructions or data structures and capable of being accessed by a computer medium, but not limited to this.

Optionally, the memory 2003 is used to store the application code for executing the solution of the present application, and the execution is controlled by the processor 2001 . The processor 2001 is configured to execute the application program code stored in the memory 2003 to implement the risk prediction method provided in any embodiment of the present application.

The electronic device provided in the embodiment of the present application is applicable to any embodiment of the foregoing method, and details are not described herein again.

Embodiments of the present application provide a computer-readable storage medium, where a computer program is stored on the computer-readable storage medium, and when the program is executed by a processor, the risk prediction method shown in the foregoing method embodiments is implemented.

The computer-readable storage medium provided by the embodiment of the present application is applicable to any embodiment of the foregoing method, and details are not described herein again.

In the risk prediction scheme provided by the embodiment of the present application, the transaction information of the target user is collected; the risk influencing factor information is determined according to the transaction information of the target user; the risk influencing factor information is input into the risk prediction model; wherein the risk prediction model is loaded A convolutional neural network model based on a swarm intelligence optimization algorithm; the risk prediction model outputs a risk prediction result, wherein the risk prediction result is used to characterize whether the target user is a risk user. The risk prediction scheme disclosed in the embodiment of this application uses the convolutional neural network model of the improved bird flock optimization algorithm to process and predict user transaction data, which can quickly, efficiently and accurately predict whether the user is a risk user, and has a high accuracy rate and a false alarm rate. low cost and low cost. Avoid risky users, such as black-produced users, who use marketing plan loopholes to arbitrage to bring unnecessary losses to merchants or customers.

It should be understood that although the various steps in the flowchart of the accompanying drawings are sequentially shown in the order indicated by the arrows, these steps are not necessarily executed in sequence in the order indicated by the arrows. Unless explicitly stated herein, the execution of these steps is not strictly limited to the order and may be performed in other orders. Moreover, at least a part of the steps in the flowchart of the accompanying drawings may include multiple sub-steps or multiple stages, and these sub-steps or stages are not necessarily executed at the same time, but may be executed at different times, and the execution sequence is also It does not have to be performed sequentially, but may be performed alternately or alternately with other steps or at least a portion of sub-steps or stages of other steps.

The above are only some embodiments of the present invention. It should be pointed out that for those skilled in the art, some improvements and modifications can be made without departing from the principles of the present invention. These improvements and modifications should also be regarded as It is the protection scope of the present invention.

Claims (12)

1. A method of risk prediction, the method comprising:

collecting target user transaction information;

determining risk influence factor information according to the target user transaction information;

inputting the risk influence factor information into a risk prediction model; the risk prediction model is a convolutional neural network model loaded with a group intelligent optimization algorithm;

and outputting a risk prediction result by a risk prediction model, wherein the risk prediction result is used for representing whether the target user is a risk user.

2. The risk prediction method of claim 1, wherein the swarm intelligence optimization algorithm is a bird swarm optimization algorithm.

3. The risk prediction method of claim 2, wherein the constructing of the risk prediction model comprises:

loading a bird group optimization algorithm into the convolutional neural network model, and determining parameters of the convolutional neural network model according to the bird group optimization algorithm;

acquiring sample transaction information to train the neural network model loaded with the bird swarm optimization algorithm;

and determining the trained neural network model loaded with the bird swarm optimization algorithm as a risk prediction model according to a preset training rule.

4. The risk prediction method of claim 3, wherein the bird swarm optimization algorithm is a bird swarm optimization algorithm that adds a boundary constraint.

5. The risk prediction method according to claim 3 or 4, wherein the preset rules comprise at least one of:

determining the convolutional neural network model loaded with the bird swarm optimization algorithm as a risk prediction model when the training iteration number reaches a preset threshold value;

or the like, or, alternatively,

inputting sample transaction data to a convolutional neural network model loaded with a bird swarm optimization algorithm, and calculating a sample fitness value after iterative training for a preset number of times;

and if the sample fitness value accords with a preset fitness value, stopping training, and determining the convolutional neural network model loaded with the bird swarm optimization algorithm as a risk prediction model.

6. The risk prediction method according to claims 1-4, characterized in that the method comprises: and collecting transaction data of the target user in real time.

7. A risk prediction device, the device comprising: the device comprises an acquisition module, a determination module, an input module, a prediction module and a storage module, wherein:

the acquisition module is used for acquiring the transaction information of the target user;

the determining module is used for determining risk influence factor information according to the target user transaction information;

the input module is used for inputting the risk influence factor information into a risk prediction model;

the storage module is used for storing a risk prediction model, wherein the risk prediction model is a convolutional neural network model loaded with a group intelligent optimization algorithm;

and the prediction module is used for controlling a risk prediction model to output a risk prediction result, wherein the risk prediction result is used for representing whether the target user is a risk user.

8. The risk prediction device of claim 7, wherein the swarm intelligence optimization algorithm is a bird swarm optimization algorithm that adds boundary constraints.

9. The risk prediction device of claim 8, further comprising a model building module, wherein the model building module is configured to:

loading a bird group optimization algorithm into the convolutional neural network model, and determining parameters of the convolutional neural network model according to the bird group optimization algorithm;

acquiring sample transaction information to train the neural network model loaded with the bird swarm optimization algorithm;

and determining the trained neural network model loaded with the bird swarm optimization algorithm as a risk prediction model according to a preset training rule.

10. The risk prediction device of claims 7-9, wherein the collection module is further configured to collect target user transaction data in real-time.

11. An electronic device comprising a processor and a memory;

the memory is used for storing operation instructions;

the processor is used for executing the method of any one of claims 1-6 by calling the operation instruction.

12. A computer-readable storage medium, characterized in that the storage medium has stored thereon a computer program which, when being executed by a processor, carries out the method of any one of claims 1-6.

Images