Transaction object risk level determination method and device, equipment, storage medium and program product

Abstract

The invention provides a transaction object risk level determination method which can be applied to the technical field of big data. The transaction object risk level determination method comprises the following steps: acquiring a plurality of transaction data of a plurality of transaction objects; performing feature coding on the plurality of transaction data to obtain a first feature subset of each transaction object; performing risk assessment on the plurality of first feature subsets, and screening out a second feature subset of each transaction object from the plurality of first feature subsets according to a risk assessment result; performing non-dominated sorting and diversification processing on the plurality of second feature subsets to obtain an initial risk feature set of each transaction object; screening the plurality of initial risk feature sets by using a genetic algorithm to obtain a target risk feature set of each transaction object; and determining the risk level of each transaction object according to the target risk feature set. The invention further provides a transaction object risk level determination device and equipment, a storage medium and a program product.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of big data, and more particularly to a transaction object risk level determination method and device, equipment, a storage medium and a program product. BACKGROUND

[0002] With the rapid development of Internet finance and e-commerce, risk assessment of transaction objects has become a core requirement of financial institutions and platform enterprises. Traditional risk level determination methods mainly rely on manual review, static rules or simple scoring card models, which are inefficient and have poor adaptability to static rules.

[0003] In addition, feature selection is also needed for risk assessment. Existing feature selection methods (such as particle swarm optimization and filtering method) have inherent defects such as single-objective optimization and static weight, and are difficult to cope with challenges such as high-dimensional sparse data, multi-objective conflict constraints and dynamic environmental changes.

[0004] Therefore, in the transaction process, how to assess the risk level of transaction objects through transaction data of transaction objects is an important way to avoid losses of users and platforms. SUMMARY

[0005] In view of the above problems, the present application provides a transaction object risk level determination method and device, equipment, a storage medium and a program product, which integrates multi-dimensional data such as transaction data, financial information and market dynamics of transaction objects, uses data analysis, modeling and monitoring technology, and identifies and assesses risks in advance to obtain the risk level of transaction objects.

[0006] According to a first aspect of the present application, a transaction object risk level determination method is provided, comprising: obtaining a plurality of transaction data of a plurality of transaction objects; performing feature encoding on the plurality of transaction data to obtain a first feature subset of each transaction object; performing risk assessment on the plurality of first feature subsets, and selecting a second feature subset of each transaction object from the plurality of first feature subsets according to the risk assessment result; performing non-dominated sorting and diversification processing on the plurality of second feature subsets to obtain an initial risk feature set of each transaction object; using a genetic algorithm to filter the plurality of initial risk feature sets to obtain a target risk feature set of each transaction object; and determining a risk level of each transaction object according to the target risk feature set.

[0007] According to an embodiment of the present application, the feature encoding of the plurality of transaction data to obtain the first feature subset of each transaction object comprises: performing feature extraction on the plurality of transaction data of each transaction object; and performing binary chromosome encoding on the plurality of transaction data after feature extraction to obtain the first feature subset of each transaction object.

[0008] According to an embodiment of the present application, the risk assessment is performed on the plurality of first feature subsets, and the second feature subset of each transaction object is selected from the plurality of first feature subsets according to the risk assessment result, including: obtaining a plurality of fitness values of each first feature subset by using a neural network model; performing risk assessment on each first feature subset by using the plurality of fitness values to obtain a risk assessment result of each first feature subset; and retaining a first feature subset with a risk assessment result higher than a preset threshold value in the plurality of first feature subsets as the second feature subset of each transaction object.

[0009] According to an embodiment of the present application, the plurality of fitness values of each first feature subset are obtained by using the neural network model, including: constructing an initial neural network model, performing feature selection on the plurality of first feature subsets by using the initial neural network model; performing standardization processing on the plurality of first feature subsets after the feature selection; updating the initial neural network model, and processing the plurality of first feature subsets after the standardization processing by using the updated initial neural network model to obtain the plurality of fitness values of each first feature subset.

[0010] According to an embodiment of the present application, the non-dominated sorting is performed on the plurality of second feature subsets, including: selecting two second feature subsets in the plurality of second feature subsets as a first solution and a second solution for non-dominated sorting; obtaining a plurality of first fitness values of the first solution and a plurality of second fitness values of the second solution by using the updated neural network model; and determining a non-dominated sorting rank of the first solution and the second solution by using the plurality of first fitness values and the plurality of second fitness values.

[0011] According to an embodiment of the present application, the non-dominated sorting rank of the first solution and the second solution is determined by using the plurality of first fitness values and the plurality of second fitness values, including: when none of the plurality of first fitness values is smaller than the plurality of second fitness values, and at least one first fitness value in the plurality of first fitness values is greater than one second fitness value in the plurality of second fitness values, the non-dominated sorting rank of the first solution is higher than that of the second solution, and the first solution is retained.

[0012] According to an embodiment of the present application, the plurality of second feature subsets are diversified to obtain an initial risk feature set of each transaction object, including: calculating a crowding degree of the plurality of first solutions retained in the plurality of second feature subsets, and retaining a plurality of first solutions with a crowding degree higher than a preset crowding value as the initial risk feature set of each transaction object.

[0013] According to an embodiment of the present application, the plurality of initial risk feature sets are screened by using a genetic algorithm to obtain a target risk feature set of each transaction object, including: performing cross and mutation processing on the plurality of initial risk feature sets by using the genetic algorithm; and when the plurality of initial risk feature sets after the mutation processing meet a preset condition, each initial risk feature set meeting the condition is taken as the target risk feature set of each transaction object.

[0014] The second aspect of the present application provides a transaction object risk level determination apparatus, comprising: a data acquisition module configured to acquire a plurality of transaction data of a plurality of transaction objects; a data encoding module configured to encode the plurality of transaction data to obtain a first feature subset of each transaction object; a data screening module configured to perform risk assessment on the plurality of first feature subsets, and screen a second feature subset of each transaction object from the plurality of first feature subsets according to the risk assessment result; perform non-dominated sorting and diversification processing on the plurality of second feature subsets to obtain an initial risk feature set of each transaction object; a screening update module configured to screen the plurality of initial risk feature sets using a genetic algorithm to obtain a target risk feature set of each transaction object; and a level determination module configured to determine a risk level of each transaction object according to the target risk feature set.

[0015] The third aspect of the present application provides an electronic device, comprising: one or more processors; a memory configured to store one or more computer programs, wherein the one or more processors execute the one or more computer programs to implement the steps of the method.

[0016] The fourth aspect of the present application further provides a computer-readable storage medium having a computer program or instructions stored thereon, wherein the computer program or instructions are executed by a processor to implement the steps of the method.

[0017] The fifth aspect of the present application further provides a computer program product comprising a computer program or instructions, wherein the computer program or instructions are executed by a processor to implement the steps of the method. BRIEF DESCRIPTION OF DRAWINGS

[0018] The above and other objects, features and advantages of the present application will become more apparent from the following description of embodiments of the present application, taken in conjunction with the accompanying drawings, in which:

[0019] Figure 1 An application scenario diagram of a transaction object risk level determination method, apparatus, device, medium and program product according to embodiments of the present application is schematically shown;

[0020] Figure 2 A flowchart of a transaction object risk level determination method according to embodiments of the present application is schematically shown;

[0021] Figure 3 A flowchart of feature encoding of a plurality of transaction data according to embodiments of the present application is schematically shown;

[0022] Figure 4 A flowchart of screening of a plurality of first feature subsets according to risk assessment results according to embodiments of the present application is schematically shown;

[0023] Figure 5 A flowchart illustrating a process of non-dominant sorting and diversification processing on a plurality of second feature subsets according to an embodiment of the present application is schematically shown;

[0024] Figure 6 A flowchart illustrating a process of screening a plurality of initial risk feature sets using a genetic algorithm according to an embodiment of the present application is schematically shown;

[0025] Figure 7 A structural block diagram of a transaction object risk level determination apparatus according to an embodiment of the present application is schematically shown; and

[0026] Figure 8 A block diagram of an electronic device adapted to implement a transaction object risk level determination method according to an embodiment of the present application is schematically shown. DETAILED DESCRIPTION

[0027] Hereinafter, embodiments of the present application will be described with reference to the accompanying drawings. It is to be understood, however, that the description is merely illustrative of the present application and in no way limits the scope of the present application. In the following detailed description, numerous specific details are set forth in order to provide a thorough understanding of the present application. However, it will be apparent to one skilled in the art that one or more embodiments can be practiced without these specific details. In other instances, well-known structures and functions have not been described in detail in order to avoid obscuring aspects of the present application.

[0028] The terms used herein are merely used to describe specific embodiments and are not intended to limit the present application. The terms "include", "comprise" and the like used herein indicate the presence of the described features, steps, operations and / or components, but do not exclude the presence or addition of one or more other features, steps, operations or components.

[0029] All terms used herein, including technical and scientific terms, have the same meanings as those generally understood by those skilled in the art unless otherwise defined. It should be noted that the terms used herein should be interpreted as having meanings consistent with the context of the present specification, and should not be interpreted in an idealized or excessively formal manner.

[0030] In the case of using expressions similar to "at least one of A, B and C, etc.", it should generally be interpreted to include at least one of A, B or C, or any combination thereof, unless otherwise defined (for example, "a system having at least one of A, B and C" should include a system having A alone, a system having B alone, a system having C alone, a system having A and B together, a system having A and C together, a system having B and C together, and / or a system having A, B, and C together, etc.).

[0031] In the technical solutions of the present application, the user information (including but not limited to user personal information, user image information, user device information such as location information, etc.) and data (including but not limited to data for analysis, stored data, displayed data, etc.) involved are all information and data authorized by the user or authorized by all parties, and the collection, storage, use, processing, transmission, provision, disclosure and application of related data comply with relevant laws, regulations and standards, take necessary security measures, do not violate public order and good customs, and provide corresponding operation portal for user selection authorization or refusal.

[0032] In the scenario of making automated decisions with personal information, the method, device and system provided by the embodiments of the present application all provide corresponding operation portal for the user to select to agree or refuse the automated decision result; if the user chooses to refuse, the expert decision process is entered. The expression "automated decision" here refers to the activity of automatically analyzing, evaluating the personal behavior habits, interests and hobbies, or economic, health, credit status, etc. by computer program and making decisions. The expression "expert decision" here refers to the activity of making decisions by personnel who are engaged in a certain field of work, have special experience, knowledge and skills and reach a certain professional level.

[0033] Figure 1 The application scenario of the transaction object risk level determination method, apparatus, device, medium and program product according to the embodiments of the present application is schematically shown.

[0034] As shown in Figure 1 The application scenario 100 according to the embodiments can include a first terminal device 101, a second terminal device 102, a third terminal device 103, a network 104 and a server 105. The network 104 is a medium for providing communication links between the first terminal device 101, the second terminal device 102, the third terminal device 103 and the server 105. The network 104 can include various connection types, such as wired, wireless communication links or optical fiber cables, etc.

[0035] The user can use the first terminal device 101, the second terminal device 102, the third terminal device 103 to interact with the server 105 through the network 104 to receive or send messages, etc. Various communication client applications can be installed on the first terminal device 101, the second terminal device 102, the third terminal device 103, such as shopping applications, web browser applications, search applications, instant messaging tools, email clients, social platform software, etc. (only as examples).

[0036] The first terminal device 101, the second terminal device 102, and the third terminal device 103 can be various electronic devices with a display screen and supporting web browsing, including but not limited to a smart phone, a tablet computer, a laptop computer, a desktop computer, and the like.

[0037] The server 105 can be a server providing various services, for example, a background management server supporting a website browsed by a user using the first terminal device 101, the second terminal device 102, and the third terminal device 103 (only as an example). The background management server can perform analysis and the like on received user requests and the like, and feed back a processing result (for example, a webpage, information, or data, or the like, obtained or generated according to a user request) to a terminal device.

[0038] It should be noted that the business view processing method provided in the embodiments of the present application can generally be executed by the server 105. Correspondingly, the business view processing apparatus provided in the embodiments of the present application can generally be arranged in the server 105. The business view processing method provided in the embodiments of the present application can also be executed by a server or a server cluster different from the server 105 and capable of communicating with the first terminal device 101, the second terminal device 102, the third terminal device 103, and / or the server 105. Correspondingly, the business view processing apparatus provided in the embodiments of the present application can also be arranged in a server or a server cluster different from the server 105 and capable of communicating with the first terminal device 101, the second terminal device 102, the third terminal device 103, and / or the server 105.

[0039] It should be understood that the number of terminal devices, networks, and servers in Figure 1 is only illustrative. According to the implementation needs, there can be any number of terminal devices, networks, and servers.

[0040] The following will be based on the scenario described in Figure 1 to describe the transaction object risk level determination method according to the embodiments of the present application in detail. Figures 2-6

[0041] Figure 2 An illustrative flowchart of the transaction object risk level determination method according to the embodiments of the present application is shown.

[0042] As shown in Figure 2 , the transaction object risk level determination method of this embodiment includes operation S210 to operation S260.

[0043] In operation S210, a plurality of transaction data of a plurality of transaction objects is obtained.

[0044] Multi-dimensional transaction data of a transaction object for risk assessment is collected. ​

[0045] For example, the transaction data includes financial information, market dynamics, transaction frequency, and transaction amount of the transaction object.

[0046] In operation S220, the plurality of transaction data is feature encoded to obtain a first feature subset of each transaction object.

[0047] The plurality of transaction data of each transaction object is preprocessed, and the preprocessed plurality of transaction data is feature extracted and encoded. The first feature subset can represent the behavior pattern, risk inclination and other key information of the transaction object. Through the pre-defined encoding rule, massive transaction data can be quickly processed, and the first feature subset can be cached and reused in different scenarios to avoid repeated calculation.

[0048] For example, real number encoding, integer encoding, permutation encoding, tree encoding, matrix encoding or binary encoding can be used to encode the plurality of transaction data after feature extraction.

[0049] In operation S230, the plurality of first feature subsets is risk evaluated, and a second feature subset of each transaction object is selected from the plurality of first feature subsets according to the risk evaluation result.

[0050] The most discriminative features are selected as the second feature subset of each transaction object through the risk evaluation result. The second feature subset can be dynamically adjusted according to the risk change, which can eliminate redundant or low correlation features, reduce noise interference, improve the accuracy of the second feature subset of each transaction object, and enhance the pertinence of risk identification.

[0051] For example, statistical methods or neural network models can be used to risk evaluate the plurality of first feature subsets.

[0052] In operation S240, the plurality of second feature subsets is non-dominantly sorted and diversified to obtain an initial risk feature set of each transaction object.

[0053] The plurality of second feature subsets is non-dominantly sorted, and the plurality of second feature subsets is hierarchically sorted through Pareto dominance relationship, so as to ensure that each level contains a second feature subset that is not fully dominated by other second feature subsets, avoid converging to a single optimal solution of the second feature subset, balance the optimization of features in the plurality of conflicting second feature subsets, preferentially retain solutions with high non-dominant sorting level, and allow low-level high-quality solutions to participate in evolution to prevent the algorithm from deviating to features in a second feature subset too early.

[0054] Diversifying the plurality of second feature subsets after non-dominant sorting can retain features with high information content but low correlation, ensure that the plurality of second feature subsets after non-dominant sorting can be uniformly distributed, avoid local aggregation, and ensure that the algorithm covers the plurality of second feature subsets of each transaction object.

[0055] In operation S250, the genetic algorithm is used to screen the plurality of initial risk feature sets to obtain a target risk feature set of each transaction object.

[0056] The selection, crossover and mutation operations are used to screen the plurality of initial risk feature sets to obtain a target risk feature set of each transaction object, which can automatically find an initial risk feature set with high discriminability, prevent overfitting, adapt to dynamic risk changes and support multi-objective optimization.

[0057] In operation S260, the risk level of each transaction object is determined according to the target risk feature set.

[0058] For example, the risk level is set to 1-10 levels, when the risk level of the transaction object is in 1-3 levels, it indicates that the transaction risk of the transaction object is low, when the risk level of the transaction object is in 4-6 levels, it indicates that the transaction object may have transaction risk and needs attention, and when the risk level of the transaction object is in 7-10 levels, it indicates that the transaction object has transaction risk and the transaction processing with the transaction object is avoided.

[0059] Figure 3 A flowchart of feature encoding of a plurality of transaction data according to an embodiment of the application is schematically shown.

[0060] As shown in FIG. 2, in some embodiments, the operation S220 can further include operations S221-S222. Figure 3

[0061] In operation S221, the plurality of transaction data of each transaction object is subjected to feature extraction.

[0062] The plurality of transaction data of each transaction object is subjected to data cleaning, processing of missing values, abnormal values and repeated data in the transaction data, standardization processing of numerical value type features such as transaction amount and transaction frequency to eliminate dimension influence, time sequence processing of transaction data with time attribute such as transaction date, extraction of key features according to different transaction data dimensions, including basic statistical features, behavior pattern features, time sequence dynamic features and correlation network features.

[0063] In operation S222, the plurality of transaction data after feature extraction is subjected to binary chromosome coding to obtain a first feature subset of each transaction object.

[0064] Further, the plurality of transaction data after feature extraction can be coded by binary coding.

[0065] ​Wherein, the multiple transaction data of each transaction object after feature extraction are encoded into binary chromosomes, the number of features of the multiple transaction data of each transaction object after feature extraction is set as m, the length of the chromosome is m, the gene of the i th feature of the multiple transaction data is x, when the value of the gene x of the i th feature is 0, it means that the i th feature is not selected by encoding, when the value of the gene x of the i th feature is 1, it means that the i th feature is selected by encoding, m, x and i are positive integers, and i is less than or equal to m.

[0066] Figure 4 A flowchart for screening a plurality of first feature subsets by a risk assessment result according to an embodiment of the present application is schematically shown.

[0067] As shown in Figure 4 In some embodiments, the operation S230 can further include operations S231-S233.

[0068] In operation S231, a plurality of fitness values of each first feature subset are obtained by using a neural network model.

[0069] The first feature subset of each transaction object is calculated by the neural network, that is, the quality of the feature selection of a plurality of transaction objects, the selected features are risk assessed by the neural network, and the area under the ROC curve (AUC), accuracy, and feature quantity are used as the plurality of fitness values of the first feature subset of each transaction object.

[0070] In the embodiment, an initial neural network model is constructed, the plurality of first feature subsets are selected by using the initial neural network model, the plurality of first feature subsets after feature selection are standardized, the initial neural network model is updated, the plurality of first feature subsets after standardization are processed by using the updated initial neural network model, and the plurality of fitness values of each first feature subset are obtained.

[0071] For example, the initial neural network model is constructed, including an input layer, a hidden layer and an output layer, the number of neurons of the input layer is determined according to the number n of features selected by encoding after feature extraction of the multiple transaction data of each transaction object, n is a positive integer, n is less than or equal to m, and assuming that the number of features is n, the number of input layer neurons is n.

[0072] For example, the first feature subset is preprocessed by using standardization and normalization to improve the training effect of the model, and the standardization of the first feature subset can be represented by the following formula:

[0073]

[0074] In the formula, is the standardized first feature subset, is the original data of the first feature subset, is the mean of the first feature subset, is the standard deviation.

[0075] Then, the initial neural network model is trained to optimize the weights and biases of the initial neural network model. The backpropagation algorithm can be used to update the parameters of the initial neural network model according to the gradient of the loss function with respect to the weights and biases.

[0076] For example, the loss function is defined as The loss function can be expressed by the following formula :

[0077]

[0078] In the formula, is the true label, is the predicted probability, is the number of selected features after feature extraction of the multiple transaction data of each transaction object, is the th feature of each transaction object, is less than or equal to , and are positive integers.

[0079] For example, the gradient of the loss function with respect to the weights and biases can be calculated using the backpropagation algorithm. The gradient can be calculated using the following formula:

[0080]

[0081]

[0082]

[0083] In the formula, is the gradient of the loss function with respect to the neuron output , is the gradient of the loss function with respect to the weight , is the gradient of the loss function with respect to the bias , is the gradient of the neuron output with respect to the weight , is the gradient of the neuron output bias of the gradient.

[0084] For example, according to the gradient information described above, the weights and biases are updated using an optimization algorithm for minimizing the loss function, and the gradient descent update can be calculated using the following formula:

[0085]

[0086]

[0087] wherein, is the learning rate, is the updated weight, is the updated bias.

[0088] When the loss function value converges, an updated initial neural network model is obtained, and the multiple first feature subsets after standardization are processed using the updated initial neural network model to obtain multiple adaptation values of each first feature subset.

[0089] In operation S232, risk assessment is performed on each first feature subset based on the multiple adaptation values, and a risk assessment result of each first feature subset is obtained.

[0090] By comprehensively evaluating the first feature subset based on the multiple adaptation values (such as risk probability, feature importance, and business rule matching degree), single-index deviation can be avoided, and the statistical analysis of the risk assessment result and the output of the statistical model are combined to reduce the interference of individual abnormal values or noise features in the first feature subset.

[0091] In operation S233, the first feature subsets with risk assessment results higher than a preset threshold in the multiple first feature subsets are retained as second feature subsets of each transaction object.

[0092] According to the multiple adaptation values of the first feature subsets of each transaction object, the feature weights are dynamically adjusted (such as giving higher risk scores to high-correlation features), ensuring that key risk features with risk assessment results higher than the preset threshold are preferentially retained, and the multiple features in the retained first feature subsets of each transaction object are taken as the second feature subsets of each transaction object.

[0093] Figure 5 A flowchart of performing non-dominated sorting and diversification processing on multiple second feature subsets according to an embodiment of the present application is schematically shown.

[0094] As Figure 5 shown, in some embodiments, the above operation S240 can further include operations S241-S244.

[0095] In operation S241, two second feature subsets from multiple second feature subsets are selected as the first and second solutions for non-dominated sorting.

[0096] For each transaction object, select two subsets of the second features from multiple subsets of the second features and perform non-dominated sorting, then determine the non-dominated sorting level of these two subsets.

[0097] In operation S242, multiple first fitness values ​​of the first solution and multiple second fitness values ​​of the second solution are obtained using the updated neural network model.

[0098] In operation S243, the non-dominated sorting order of the first and second solutions is determined by multiple first fitness values ​​and multiple second fitness values.

[0099] In this embodiment, when multiple first fitness values ​​are not less than multiple second fitness values, and at least one of the multiple first fitness values ​​is greater than one of the multiple second fitness values, the non-dominated ranking level of the first solution is higher than the non-dominated ranking level of the second solution, and the first solution is retained.

[0100] In operation S244, the crowding degree of multiple first solutions retained in multiple second feature subsets is calculated, and multiple first solutions with crowding degree higher than the preset crowding value are retained as the initial risk feature set of each transaction object.

[0101] For example, to retain multiple first solutions with high non-dominated ranking, the crowding degree of multiple first solutions can be calculated using the following formula:

[0102]

[0103] In the formula, d i f is the crowding degree of the i-th feature of the first solution. j Let j be the j-th target value, and n be the total number of features selected by encoding after feature extraction from multiple transaction data of each transaction object, where n > 2 and n is a positive integer.

[0104] Based on the calculated congestion level, multiple first solutions with congestion levels higher than the preset congestion value are retained as the initial risk characteristic set for each trading object.

[0105] Figure 6 The flowchart illustrating the use of a genetic algorithm to screen multiple initial risk feature sets according to an embodiment of this application is shown.

[0106] like Figure 6 As shown, in some embodiments, the above-described operation S250 may further include operations S251 to S252.

[0107] In operation S251, the genetic algorithm is used to perform cross and mutation processing on the plurality of initial risk feature sets.

[0108] Further, single-point cross can be used to process the plurality of initial risk feature sets.

[0109] For example, assuming that there are y initial risk feature sets, two initial risk feature sets x1 and x2 are selected from the y initial risk feature sets, the gene sequence of x1 is x1=x 11 ,x 12 ,……,x 1y , and the gene sequence of x2 is x2=x 21 ,x 22 ,……,x 2y , the cross point of the two initial risk feature sets is c, and after cross operation, two new initial risk feature sets are generated: child1=x 1,1 ,x 1,2 ,…,x 1,c ,x 2,c+1 ,…,x 2,y and child2=x 2,1 ,x 2,2 ,…,x 2,c ,x 1,c+1 ,…,x 1,y , and mutation processing is performed on the plurality of new initial risk feature sets.

[0110] In operation S252, when the plurality of initial risk feature sets after mutation processing meet a preset condition, each initial risk feature set meeting the condition is used as a target risk feature set of each transaction object.

[0111] The above step S251 is repeated, and when a maximum number of repetitions is reached or the plurality of fitness of the plurality of initial risk feature sets is no longer significantly improved, each initial risk feature set meeting the condition is used as a target risk feature set of each transaction object.

[0112] In summary, the transaction object risk level determination method provided by the embodiments of the present application can achieve the following beneficial effects:

[0113] (1) Through the pre-defined coding rule, massive transaction data can be quickly processed, and the feature subset is cached and reused in different scenarios, avoiding repeated calculation and improving efficiency.

[0114] (2) Through the sliding window evolution mechanism and the non-dominated sorting, online updating of the feature subset of each transaction object is realized, the optimal balance point of the multi-dimensional transaction data target of each transaction object is directly searched, and the typical conflict in the risk level determination of the transaction object is solved.

[0115] Based on the processing method of the business view, the application further provides a transaction object risk level determination apparatus. The following will be described in combination with Figure 7 The apparatus is described in detail.

[0116] Figure 7 The structure block diagram of the transaction object risk level determination apparatus according to the embodiment of the application is schematically shown.

[0117] As shown in the figure, the transaction object risk level determination apparatus 700 of the embodiment comprises a data acquisition module 710, a data coding module 720, a data screening module 730 and a level determination module 740. Figure 7 The data acquisition module 710 is used for acquiring a plurality of transaction data of a plurality of transaction objects. In an embodiment, the data acquisition module 710 can be used for performing the operation S210 described above, which will not be repeated here.

[0118] The data coding module 720 is used for performing feature coding on the plurality of transaction data to obtain a first feature subset of each transaction object. In an embodiment, the data coding module 720 can be used for performing the operation S220 described above, which will not be repeated here.

[0119] The data screening module 730 is used for performing risk assessment on the plurality of first feature subsets, screening a second feature subset of each transaction object from the plurality of first feature subsets according to the risk assessment result; performing non-dominated sorting and diversification processing on the plurality of second feature subsets to obtain an initial risk feature set of each transaction object, and screening the plurality of initial risk feature sets using a genetic algorithm to obtain a target risk feature set of each transaction object. In an embodiment, the data screening module 730 can be used for performing the operations S230-S250 described above, which will not be repeated here.

[0120] The level determination module 740 is used for determining a risk level of each transaction object according to the target risk feature set. In an embodiment, the level determination module 740 can be used for performing the operation 260 described above, which will not be repeated here.

[0121] According to the embodiment of the application, the data coding module 720 comprises a feature extraction unit and a data coding unit.

[0122] The feature extraction unit is used for performing feature extraction on the plurality of transaction data of each transaction object.

[0123] The data coding unit is used for performing binary chromosome coding on the plurality of transaction data after feature extraction to obtain the first feature subset of each transaction object.

[0124]

[0125] ​According to an embodiment of the present application, the data screening module 730 comprises an adaptive value obtaining unit, a risk assessment unit and a subset screening unit.

[0126] The adaptive value obtaining unit is configured to obtain a plurality of adaptive values of each first feature subset by using the neural network model.

[0127] The risk assessment unit is configured to perform risk assessment on each first feature subset by using the plurality of adaptive values, to obtain a risk assessment result of each first feature subset.

[0128] The subset screening unit is configured to retain a first feature subset with a risk assessment result higher than a preset threshold in the plurality of first feature subsets as a second feature subset of each transaction object.

[0129] According to an embodiment of the present application, the adaptive value obtaining unit further comprises a feature selection subunit, a subset processing subunit and an adaptive value obtaining subunit.

[0130] The feature selection subunit is configured to construct an initial neural network model, and perform feature selection on the plurality of first feature subsets by using the initial neural network model.

[0131] The subset processing subunit is configured to perform standardization processing on the plurality of first feature subsets after feature selection.

[0132] The adaptive value obtaining subunit is configured to update the initial neural network model, and perform processing on the plurality of first feature subsets after standardization processing by using the updated initial neural network model, to obtain a plurality of adaptive values of each first feature subset.

[0133] According to an embodiment of the present application, the data screening module 730 further comprises a solution set selection unit, a solution value obtaining unit and a rank determination unit.

[0134] The solution set selection unit is configured to select two second feature subsets in the plurality of second feature subsets as a first solution and a second solution of non-dominated sorting.

[0135] The solution value obtaining unit is configured to obtain a plurality of first adaptive values of the first solution and a plurality of second adaptive values of the second solution by using the updated neural network model.

[0136] The rank determination unit is configured to determine a non-dominated sorting rank of the first solution and the second solution by using the plurality of first adaptive values and the plurality of second adaptive values.

[0137] According to an embodiment of the present application, the rank determination unit is further configured to, when none of the plurality of first adaptive values is smaller than the plurality of second adaptive values, and at least one first adaptive value in the plurality of first adaptive values is greater than one second adaptive value in the plurality of second adaptive values, the non-dominated sorting rank of the first solution is higher than the non-dominated sorting rank of the second solution, and the first solution is retained.

[0138] According to an embodiment of the present application, the data screening module 730 further comprises a crowdedness determination unit.

[0139] The crowdedness determination unit is configured to calculate crowdedness of the plurality of first solutions reserved in the plurality of second feature subsets, and reserve the plurality of first solutions with crowdedness higher than a preset crowdedness value as the initial risk feature set of each transaction object.

[0140] According to an embodiment of the present application, the data screening module 730 further comprises a crossover and mutation unit, and a risk set unit.

[0141] The crossover and mutation unit is configured to perform crossover and mutation processing on the plurality of initial risk feature sets using a genetic algorithm.

[0142] The risk set unit is configured to take each initial risk feature set satisfying a preset condition as the target risk feature set of each transaction object when the plurality of initial risk feature sets after mutation processing satisfy the preset condition.

[0143] According to an embodiment of the present application, any of the data acquisition module 710, the data encoding module 720, the data screening module 730 and the level determination module 740 can be combined in one module for implementation, or any of the modules can be split into multiple modules. Alternatively, at least part of the function of one or more of the modules can be combined with at least part of the function of other modules, and implemented in one module. According to an embodiment of the present application, at least one of the data acquisition module 710, the data encoding module 720, the data screening module 730 and the level determination module 740 can be at least partially implemented as a hardware circuit, such as a field programmable gate array (FPGA), a programmable logic array (PLA), a system on chip, a system on board, a system on package, an application specific integrated circuit (ASIC), or any other reasonable manner of hardware or firmware that can be integrated or packaged with a circuit, or implemented in any one of software, hardware and firmware or in an appropriate combination of any of them. Alternatively, at least one of the data acquisition module 710, the data encoding module 720, the data screening module 730 and the level determination module 740 can be at least partially implemented as a computer program module that can perform corresponding functions when the computer program module is run.

[0144] Figure 8 A block diagram of an electronic device suitable for implementing determination of the risk level of a transaction object according to an embodiment of the present application is schematically shown.

[0145] As Figure 8As shown, the electronic device 800 according to embodiments of the present application includes a processor 801 which can perform various appropriate actions and processes in accordance with a program stored in a read only memory (ROM) 802 or a program loaded into a random access memory (RAM) 803 from a storage section 808. The processor 801 can include, for example, a general purpose microprocessor (e.g., a CPU), an instruction set processor, and / or a related chip set, and / or a dedicated microprocessor (e.g., an application specific integrated circuit (ASIC)), and so on. The processor 801 can also include an on-board memory for cache use. The processor 801 can include a single processing unit or multiple processing units for executing different actions of the method processes according to embodiments of the present application.

[0146] In the RAM 803, various programs and data required for the operation of the electronic device 800 are stored. The processor 801, the ROM 802, and the RAM 803 are connected to each other via a bus 804. The processor 801 performs various operations of the method processes according to embodiments of the present application by executing the programs in the ROM 802 and / or the RAM 803. Note that the programs can also be stored in one or more memories other than the ROM 802 and the RAM 803. The processor 801 can also perform various operations of the method processes according to embodiments of the present application by executing the programs stored in the one or more memories.

[0147] According to embodiments of the present application, the electronic device 800 can also include an input / output (I / O) interface 805 which is also connected to the bus 804. The electronic device 800 can also include one or more of the following components connected to the input / output (I / O) interface 805: an input section 808 including a keyboard, a mouse, etc.; an output section 807 including a display such as a cathode ray tube (CRT), a liquid crystal display (LCD), etc., and a speaker, etc.; a storage section 808 including a hard disk, etc.; and a communication section 809 including a network interface card such as a LAN card, a modem, etc. The communication section 809 performs communication processing via a network such as the Internet. A drive 810 is also connected to the input / output (I / O) interface 805 as necessary. A removable recording medium 811 such as a magnetic disk, an optical disk, a magneto-optical disk, a semiconductor memory, etc. is attached to the drive 810 as necessary, so that a computer program read out therefrom is installed in the storage section 808 as necessary.

[0148] The application further provides a computer readable storage medium, which can be included in the device / apparatus / system described in the above embodiments, or can exist independently without being assembled into the device / apparatus / system. The computer readable storage medium carries one or more programs, which, when executed, implement the method according to the embodiments of the application.

[0149] According to the embodiments of the application, the computer readable storage medium can be a non-volatile computer readable storage medium, which can include, but is not limited to, a portable computer diskette, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), a portable compact disc read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any appropriate combination thereof. In this application, a computer readable storage medium can be any tangible medium that contains or stores a program for use by or in connection with an instruction execution system, apparatus, or device. For example, according to the embodiments of the application, the computer readable storage medium can include the ROM 802 and / or the RAM 803 described above, and / or one or more memory other than the ROM 802 and the RAM 803.

[0150] The embodiments of the application also include a computer program product, which includes a computer program containing program codes for executing the method shown in the flow chart. When the computer program product is run in a computer system, the program codes are used to make the computer system implement the processing method of the service view provided by the embodiments of the application.

[0151] The above functions defined in the system / apparatus of the embodiments of the application are performed when the computer program is executed by the processor 801. According to the embodiments of the application, the system, apparatus, module, unit, etc. described above can be implemented by computer program modules.

[0152] In one embodiment, the computer program can rely on a tangible storage medium such as an optical storage device, a magnetic storage device, etc. In another embodiment, the computer program can also be transmitted, distributed, and downloaded in the form of a signal on a network medium, and be downloaded and installed through the communication part 809, and / or installed from the detachable medium 811. The program codes contained in the computer program can be transmitted by any appropriate network medium, including but not limited to wireless, wired, etc., or any appropriate combination thereof.

[0153] In such embodiments, the computer program can be downloaded and installed from the network via the communication section 809, and / or installed from the removable media 811. When the computer program is executed by the processor 801, the above-described functions defined in the system of the embodiments of the present application are executed. According to the embodiments of the present application, the system, device, apparatus, module, unit, and the like described above can be realized by the computer program modules.

[0154] According to the embodiments of the present application, the program code for executing the computer program provided by the embodiments of the present application can be written in any combination of one or more programming languages, and specifically, these computer programs can be implemented using high-level procedural and / or object-oriented programming language, and / or assembly / machine language. The programming language includes, but is not limited to, such as Java, C++, python, "C" language, or similar programming language. The program code can be executed entirely on the user computing device, partially on the user device, partially on a remote computing device, or entirely on a remote computing device or server. In the case involving a remote computing device, the remote computing device can be connected to the user computing device through any kind of network, including a local area network (LAN) or a wide area network (WAN), or can be connected to an external computing device (for example, connected to the Internet through an Internet service provider).

[0155] The flowcharts and block diagrams in the drawings illustrate the possible architectures, functions, and operations of systems, methods, and computer program products according to various embodiments of the present application. In this regard, each block in the flowcharts or block diagrams can represent a module, a program segment, or a portion of code, which contains one or more executable instructions for implementing the specified logical functions. It should also be noted that in some alternative implementations, the functions noted in the blocks can occur in a different order than that shown in the figures. For example, two blocks noted in succession can actually be executed substantially concurrently, or they can sometimes be executed in reverse order, depending on the functionality involved. It should also be noted that each block in the flowcharts or block diagrams, and combinations of blocks in the flowcharts or block diagrams, can be implemented by dedicated hardware-based systems that perform the specified functions or operations, or can be implemented by a combination of dedicated hardware and computer instructions.

[0156] Those skilled in the art can understand that the features described in various embodiments of the present application can be combined and / or integrated in various combinations and / or integrations, even if such combinations or integrations are not explicitly described in the present application. In particular, the features described in various embodiments of the present application can be combined and / or integrated in various combinations and / or integrations without departing from the spirit and teachings of the present application. All such combinations and / or integrations fall within the scope of the present application.

Claims

1. A method of determining a risk rating of a transaction subject, characterized by, The method comprises: obtaining a plurality of transaction data of a plurality of transaction objects; feature encoding the plurality of transaction data to obtain a first feature subset of each of the transaction objects; risk assessment on a plurality of the first feature subsets, and screening a second feature subset of each of the transaction objects from the plurality of the first feature subsets according to a risk assessment result; non-dominated sorting and diversification processing on the plurality of second feature subsets to obtain an initial risk feature set of each of the transaction objects; screening a plurality of the initial risk feature sets using a genetic algorithm to obtain a target risk feature set of each of the transaction objects; determining a risk level of each of the transaction objects according to the target risk feature set.

2. The method of claim 1, wherein, The feature encoding of the plurality of transaction data to obtain a first feature subset of each of the transaction objects comprises: feature extraction of the plurality of transaction data of each of the transaction objects; binary chromosome encoding of the plurality of transaction data after feature extraction to obtain a first feature subset of each of the transaction objects.

3. The method of claim 1, wherein, The risk assessment on a plurality of the first feature subsets, and screening a second feature subset of each of the transaction objects from the plurality of the first feature subsets according to a risk assessment result comprises: obtaining a plurality of fitness values of each of the first feature subsets by using a neural network model; risk assessment on each of the first feature subsets by using the plurality of fitness values to obtain a risk assessment result of each of the first feature subsets; retaining a first feature subset with a risk assessment result higher than a preset threshold in the plurality of the first feature subsets as a second feature subset of each of the transaction objects.

4. The method of claim 3, wherein, The obtaining of a plurality of fitness values of each of the first feature subsets by using a neural network model comprises: constructing an initial neural network model, and performing feature selection on a plurality of the first feature subsets by using the initial neural network model; standardizing the plurality of first feature subsets after feature selection; updating the initial neural network model, and processing the plurality of first feature subsets after standardization by using the updated initial neural network model to obtain a plurality of fitness values of each of the first feature subsets.

5. The method of claim 4, wherein, The non-dominated sorting of the plurality of second feature subsets comprises: selecting two second feature subsets in the plurality of second feature subsets as a first solution and a second solution of the non-dominated sorting; obtaining a plurality of first fitness values of the first solution and a plurality of second fitness values of the second solution by using the updated neural network model; determining a non-dominated sorting level of the first solution and the second solution by using the plurality of first fitness values and the plurality of second fitness values.

6. The method of claim 5, wherein, The determination of a non-dominated sorting level of the first solution and the second solution by using the plurality of first fitness values and the plurality of second fitness values comprises: when none of the plurality of first fitness values is smaller than the plurality of second fitness values, and at least one of the plurality of first fitness values is greater than one of the plurality of second fitness values, the non-dominated sorting level of the first solution is higher than that of the second solution, and the first solution is retained.

7. The method of claim 6, wherein, The diversification processing on the plurality of second feature subsets obtains an initial risk feature set of each transaction object, and the initial risk feature set of each transaction object includes: The crowdedness of the plurality of first solutions retained in the plurality of second feature subsets is calculated, and the plurality of first solutions with a crowdedness higher than a preset crowdedness value are retained as the initial risk feature set of each transaction object.

8. The method of claim 1, wherein, The screening of the plurality of initial risk feature sets by using the genetic algorithm obtains a target risk feature set of each transaction object, and the target risk feature set of each transaction object includes: The plurality of initial risk feature sets are processed by using the genetic algorithm for crossover and mutation; When the plurality of initial risk feature sets after the mutation processing meet a preset condition, each initial risk feature set meeting the condition is taken as the target risk feature set of each transaction object.

9. A transaction object risk level determination apparatus characterized by comprising: The device includes: A data acquisition module is configured to acquire a plurality of transaction data of a plurality of transaction objects. A data encoding module is configured to encode the plurality of transaction data to obtain a first feature subset of each transaction object. A data screening module is configured to perform risk assessment on a plurality of first feature subsets, select a second feature subset of each transaction object from the plurality of first feature subsets according to a risk assessment result, perform non-dominated sorting and diversification processing on the plurality of second feature subsets to obtain an initial risk feature set of each transaction object, and screen a plurality of initial risk feature sets by using a genetic algorithm to obtain a target risk feature set of each transaction object. A level determination module is configured to determine a risk level of each transaction object according to the target risk feature set.

10. An electronic device, comprising: one or more processors; a memory for storing one or more computer programs, characterized in that the one or more processors execute the one or more computer programs to implement the steps of the method according to any one of claims 1-8.

11. A computer readable storage medium having stored thereon a computer program or instructions, characterized in that, The computer program or instruction is executed by the processor to implement the steps of the method according to any one of claims 1-8.

12. A computer program product comprising computer programs or instructions, characterized in that, The computer program or instruction is executed by the processor to implement the steps of the method according to any one of claims 1-8. The computer program or instruction is executed by the processor to implement the steps of the method according to any one of claims 1-8.

Images