CN114881783A - Abnormal card identification method and device, electronic equipment and storage medium - Google Patents
Abnormal card identification method and device, electronic equipment and storage medium Download PDFInfo
- Publication number
- CN114881783A CN114881783A CN202210527967.XA CN202210527967A CN114881783A CN 114881783 A CN114881783 A CN 114881783A CN 202210527967 A CN202210527967 A CN 202210527967A CN 114881783 A CN114881783 A CN 114881783A
- Authority
- CN
- China
- Prior art keywords
- card
- bank card
- bank
- abnormal
- transaction
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q40/00—Finance; Insurance; Tax strategies; Processing of corporate or income taxes
- G06Q40/04—Trading; Exchange, e.g. stocks, commodities, derivatives or currency exchange
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
- G06F18/214—Generating training patterns; Bootstrap methods, e.g. bagging or boosting
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/22—Matching criteria, e.g. proximity measures
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Data Mining & Analysis (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Evolutionary Biology (AREA)
- Evolutionary Computation (AREA)
- Bioinformatics & Computational Biology (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Artificial Intelligence (AREA)
- Life Sciences & Earth Sciences (AREA)
- Business, Economics & Management (AREA)
- Accounting & Taxation (AREA)
- Finance (AREA)
- Development Economics (AREA)
- Economics (AREA)
- Marketing (AREA)
- Strategic Management (AREA)
- Technology Law (AREA)
- General Business, Economics & Management (AREA)
- Financial Or Insurance-Related Operations Such As Payment And Settlement (AREA)
Abstract
The invention discloses an abnormal card identification method, an abnormal card identification device, electronic equipment and a storage medium, aiming at the same kind of bank cards, after determining a characteristic vector of the bank card according to a historical transaction record of the bank card and determining a probability value of the bank card as the abnormal card based on an abnormal card identification model, determining the target similarity of the bank card and the same kind of transacted bank cards, counting the accumulated transaction number and the transaction interval time of the same kind of transacted bank cards, and further determining a risk exposure parameter value according to the probability value, the target similarity, the accumulated transaction number and the transaction interval time. And finally, performing abnormal card identification based on the risk exposure parameter value. Firstly, determining the probability value of the bank card as an abnormal card, and then jointly determining the risk exposure parameter value for identifying the abnormal behavior by combining the tendency of the bank card, namely the target similarity, the accumulated transaction number and the transaction interval time. The method makes up the singleness of the existing rule and improves the accuracy of abnormal card identification.
Description
Technical Field
The invention relates to the technical field of internet finance, in particular to an abnormal card identification method and device, electronic equipment and a storage medium.
Background
The activity discount in the enterprise marketing process is beneficial to the enterprise to refresh and improve the liveness and brand image of users, but some people try to obtain benefits by carrying out abnormal transactions through marketing activities, and the behaviors are called abnormal behaviors. The abnormal behaviors seriously affect the marketing cost of enterprises and reduce the marketing effect. Therefore, how to detect and handle these abnormal behaviors and ensure that marketing funds flow to consumers is an important guarantee for enterprises to develop marketing activities.
Generally, the abnormal cards have the characteristics of identical card Bin (bank identification code), certain continuity of card numbers, convergence of transaction modes and the like. The currently adopted rule decision technology judges rule strategies by loading corresponding abnormal cards, if more than three cards with the same Bin are found, early warning is carried out, and the judgment results are given and then checked by business personnel. However, the rule decision technology is excessively dependent on expert experience, is limited to the identification of continuous card numbers, and has a single design framework, so that the risk of interfering normal transaction is high, and the enthusiasm of normal consumers for participating in marketing activities is struck. When similar alarms are processed by the market department, an effective auxiliary verification means is lacked, so that customer complaints are particularly high, the opinion of a merchant is large, and the market popularization is influenced. In view of this, it is difficult for such methods to meet the increasingly complex requirements of abnormal card identification scenarios.
Disclosure of Invention
The embodiment of the invention provides an abnormal card identification method, an abnormal card identification device, electronic equipment and a storage medium, which are used for solving the problem of high error identification rate of the existing abnormal card identification scheme.
The embodiment of the invention provides an abnormal card identification method, which comprises the following steps:
classifying the bank cards in which the transaction occurs by taking the bank identification codes as dimensions;
for each type of bank card, determining a feature vector of the bank card according to a historical transaction record of the bank card, inputting the feature vector of the bank card into a pre-trained abnormal card identification model, and determining a probability value of the bank card as an abnormal card based on the abnormal card identification model;
determining the target similarity of the bank card and the similar transacted bank card according to the characteristic vectors corresponding to the bank card and the similar transacted bank card;
counting the accumulated transaction number and the transaction interval time of the same type of transacted bank card, and determining a risk exposure parameter value according to the probability value, the target similarity, the accumulated transaction number and the transaction interval time;
and performing abnormal card identification on the bank card according to the risk exposure parameter value.
Further, determining the feature vector of the bank card according to the historical transaction record of the bank card comprises:
respectively determining a basic characteristic vector, a transaction characteristic vector and a transaction network characteristic vector of the bank card according to the historical transaction record of the bank card;
and determining the characteristic vector of the bank card according to the basic characteristic vector, the transaction characteristic vector and the transaction network characteristic vector of the bank card.
Further, the training process of the abnormal card identification model comprises the following steps:
and aiming at each sample bank card in the training set, inputting the sample characteristic vector of the sample bank card and the corresponding marking information of whether the sample bank card is an abnormal card into the abnormal card identification model, and training the abnormal card identification model.
Further, determining the target similarity between the bank card and the similar transacted bank card according to the feature vectors corresponding to the bank card and the similar transacted bank card comprises:
respectively determining each similarity of the bank card and the similar transacted bank card according to the characteristic vectors corresponding to the bank card and the similar transacted bank card;
and selecting the maximum similarity from the similarity as the target similarity.
Further, the determining a risk exposure parameter value according to the probability value, the target similarity, the accumulated transaction number and the transaction interval time comprises:
substituting the probability value, the target similarity, the accumulated transaction stroke number and the transaction interval time into the following formula to determine a risk exposure parameter value;
in the formula, G t And f (X) is the probability value that the bank card is an abnormal card, alpha is a hyper-parameter for adjusting the weight, S is the target similarity, N is the accumulated transaction number, and T is the transaction interval time.
Further, the performing abnormal card identification on the bank card according to the risk exposure parameter value comprises:
and judging whether the risk exposure parameter value is larger than a preset risk threshold value, if so, determining that the bank card is an abnormal card, and if not, determining that the bank card is a normal card.
Further, the method further comprises:
and if the bank card is determined to be an abnormal card, determining all the transacted bank cards similar to the bank card to be abnormal cards.
On the other hand, an embodiment of the present invention provides an abnormal card identification apparatus, including:
the classification module is used for classifying the bank cards in which the transaction occurs by taking the bank identification codes as dimensions;
the first determining module is used for determining a feature vector of each type of bank card according to the historical transaction record of the bank card, inputting the feature vector of the bank card into a pre-trained abnormal card identification model, and determining the probability value of the bank card as the abnormal card based on the abnormal card identification model;
the second determining module is used for determining the target similarity of the bank card and the similar transacted bank card according to the characteristic vectors corresponding to the bank card and the similar transacted bank card;
the third determining module is used for counting the accumulated transaction stroke number and the transaction interval time of the similar transacted bank cards and determining a risk exposure parameter value according to the probability value, the target similarity, the accumulated transaction stroke number and the transaction interval time;
and the identification module is used for carrying out abnormal card identification on the bank card according to the risk exposure parameter value.
Further, the first determining module is specifically configured to determine a basic feature vector, a transaction feature vector, and a transaction network feature vector of the bank card according to a historical transaction record of the bank card; and determining the characteristic vector of the bank card according to the basic characteristic vector, the transaction characteristic vector and the transaction network characteristic vector of the bank card.
Further, the apparatus further comprises:
and the training module is used for inputting the sample characteristic vectors of the sample bank cards and the corresponding marking information of whether the sample bank cards are abnormal cards or not into the abnormal card identification model aiming at each sample bank card in the training set, and training the abnormal card identification model.
Further, the second determining module is specifically configured to determine, according to the feature vectors corresponding to the bank card and the like transacted bank cards, respective similarities between the bank card and the like transacted bank cards respectively; and selecting the maximum similarity from the similarity as the target similarity.
Further, the third determining module is specifically configured to substitute the probability value, the target similarity, the accumulated transaction stroke number, and the transaction interval time into the following formula to determineDetermining a risk exposure parameter value;
in the formula, G t And f (X) is a probability value that the bank card is an abnormal card, alpha is a hyperparameter for adjusting the weight, S is the target similarity, N is the accumulated transaction number, and T is the transaction interval time.
Further, the identification module is specifically configured to determine whether the risk exposure parameter value is greater than a preset risk threshold, determine that the bank card is an abnormal card if the risk exposure parameter value is greater than the preset risk threshold, and determine that the bank card is a non-abnormal card if the risk exposure parameter value is not greater than the preset risk threshold.
Further, the identification module is further configured to determine that all the transacted bank cards similar to the bank card are abnormal cards if the bank card is determined to be an abnormal card.
In another aspect, an embodiment of the present invention provides an electronic device, including a processor, a communication interface, a memory, and a communication bus, where the processor, the communication interface, and the memory complete mutual communication through the communication bus;
a memory for storing a computer program;
a processor for implementing any of the above method steps when executing a program stored in the memory.
In yet another aspect, an embodiment of the present invention provides a computer-readable storage medium, in which a computer program is stored, and the computer program, when executed by a processor, implements the method steps of any one of the above.
The embodiment of the invention provides an abnormal card identification method, an abnormal card identification device, electronic equipment and a storage medium, wherein the method comprises the following steps: classifying the bank cards in which the transaction occurs by taking the bank identification code as a dimension; for each type of bank card, determining a feature vector of the bank card according to a historical transaction record of the bank card, inputting the feature vector of the bank card into a pre-trained abnormal card identification model, and determining a probability value of the bank card as an abnormal card based on the abnormal card identification model; determining the target similarity of the bank card and the similar transacted bank card according to the characteristic vectors corresponding to the bank card and the similar transacted bank card; counting the accumulated transaction number and the transaction interval time of the same type of transacted bank card, and determining a risk exposure parameter value according to the probability value, the target similarity, the accumulated transaction number and the transaction interval time; and performing abnormal card identification on the bank card according to the risk exposure parameter value.
The technical scheme has the following advantages or beneficial effects:
in the embodiment of the invention, for the same kind of bank cards, after the characteristic vector of the bank card is determined according to the historical transaction record of the bank card, the probability value of the bank card as the abnormal card is determined based on the abnormal card identification model, the target similarity of the bank card and the similar transacted bank cards is determined, the accumulated transaction number and the transaction interval time of the similar transacted bank cards are counted, and the risk exposure parameter value is determined according to the probability value, the target similarity, the accumulated transaction number and the transaction interval time. And finally, performing abnormal card identification based on the risk exposure parameter value. The embodiment of the invention firstly determines the probability value of the bank card as the abnormal card by utilizing the characteristic vector of the bank card, and then jointly determines the risk exposure parameter value for identifying the abnormal behavior by combining the tendency of the bank card, namely the target similarity, and the accumulated transaction number and the transaction interval time. The method makes up the singleness of the existing rule and improves the accuracy of abnormal card identification.
Drawings
In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the drawings needed to be used in the description of the embodiments will be briefly introduced below, and it is obvious that the drawings in the following description are only some embodiments of the present invention, and it is obvious for those skilled in the art to obtain other drawings based on these drawings without creative efforts.
Fig. 1 is a schematic diagram of an abnormal card identification process according to an embodiment of the present invention;
FIG. 2 is a flow chart of the exception card identification provided by the embodiment of the present invention;
FIG. 3 is a diagram of an abnormal card recognition framework based on supervised learning algorithm according to an embodiment of the present invention;
FIG. 4 is a flowchart illustrating an overall implementation of the abnormal card identification according to an embodiment of the present invention;
fig. 5 is a schematic structural diagram of an abnormal card identification apparatus according to an embodiment of the present invention;
fig. 6 is a schematic structural diagram of an electronic device according to an embodiment of the present invention.
Detailed Description
The present invention will be described in further detail with reference to the attached drawings, and it should be understood that the described embodiments are only a part of the embodiments of the present invention, and not all of the embodiments. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.
The abnormal card identification method of the related art has the following disadvantages:
(1) corresponding expert rules are set through historical experience, such as the number of cards for identifying the same card bin (bank identification code), the detection logic is single, and the risk information of the card holders and the information mining of transaction behavior sequences in marketing activities are insufficient.
(2) The method mainly aims at post detection, has insufficient generalization capability, does not take the marketing activities as a system, considers the dynamics of abnormal card risks, and cannot effectively block the abnormal behaviors in the occurrence process.
(3) The method is very easy to interfere normal cardholders to participate in marketing activities and influence user experience, so that the influence of the marketing activities is reduced, and the conventional mode is not suitable for a real-time abnormal card detection system.
Aiming at the defects, the invention innovatively provides an abnormal card detection method based on a dynamic risk exposure algorithm. Firstly, obtaining a probability value of a bank card as an abnormal card by adopting an abnormal card identification model, and solving the problems of dependence on expert experience and incomplete comprehensiveness; secondly, dynamically measuring the risk exposure coefficients of a plurality of cardholders with the same Bin participating in marketing activities by modeling the transaction behavior sequence of the cardholders; and finally, a detection strategy in the affairs is provided, so that the abnormal card can be identified, meanwhile, the interference to normal transaction can be reduced as far as possible on the premise of protecting the safety of marketing activities, and the marketing experience of users can be improved.
Fig. 1 is a schematic diagram of an abnormal card identification process provided in an embodiment of the present invention, where the process includes the following steps:
s101: and classifying the bank cards in which the transaction occurs by taking the bank identification codes as dimensions.
S102: for each type of bank card, determining a feature vector of the bank card according to a historical transaction record of the bank card, inputting the feature vector of the bank card into a pre-trained abnormal card identification model, and determining a probability value of the bank card as an abnormal card based on the abnormal card identification model.
S103: and determining the target similarity of the bank card and the bank card which is transacted in the same type according to the characteristic vectors which are respectively corresponding to the bank card and the bank card which is transacted in the same type.
S104: and counting the accumulated transaction number and the transaction interval time of the bank cards which are transacted in the same category, and determining a risk exposure parameter value according to the probability value, the target similarity, the accumulated transaction number and the transaction interval time.
S105: and performing abnormal card identification on the bank card according to the risk exposure parameter value.
In the embodiment of the invention, the bank cards which are transacted in the period of the marketing campaign are firstly classified, and the classification principle can be that the bank cards are classified according to the dimension of the bank identification code. For example, bank cards of construction banks, bank cards of industrial and commercial banks, bank cards of agricultural banks, and the like, in which transactions occur during a marketing campaign, are used as one type.
And aiming at each type of bank card, performing abnormal card identification on the type of bank card. Abnormal card identification among various bank cards can be processed in parallel. When the abnormal card identification is carried out on the bank card which is currently transacted, firstly, the historical transaction record of the bank card which is currently transacted is obtained, and then the characteristic vector of the bank card is determined according to the historical transaction distance. Specifically, determining the feature vector of the bank card according to the historical transaction record of the bank card comprises: respectively determining a basic characteristic vector, a transaction characteristic vector and a transaction network characteristic vector of the bank card according to the historical transaction record of the bank card; and determining the characteristic vector of the bank card according to the basic characteristic vector, the transaction characteristic vector and the transaction network characteristic vector of the bank card.
The basic feature vector of the bank card comprises the card type of the bank card, the bank to which the bank card belongs and the like. The card types of the bank card include credit cards, savings cards, and the like. The transaction feature vector includes historical consumption levels, historical consumption habits, and the like. The transaction network feature vector comprises a first-order neighbor node number, a second-order neighbor node number and the like. The transaction network feature vector is obtained by constructing a heterogeneous transaction network with the card holder card number and the merchant number as nodes and transaction information as edges. The number of first-order neighbor nodes refers to the number of immediately adjacent nodes, the number of second-order neighbor nodes refers to the number of nodes adjacent to the immediately adjacent nodes, and so on. Considering that the abnormal card is different from the normal card, it is mainly reflected on the above feature vector. Therefore, the feature vector needs to be determined when the abnormal card is identified.
The method comprises the following steps of pre-training and finishing an abnormal card identification model, wherein the training process of the abnormal card identification model comprises the following steps:
and aiming at each sample bank card in the training set, inputting the sample characteristic vector of the sample bank card and the corresponding marking information of whether the sample bank card is an abnormal card into the abnormal card identification model, and training the abnormal card identification model.
The training set comprises a large number of sample bank cards, and each sample bank card has corresponding labeling information whether the sample bank card is an abnormal card. When the abnormal card identification model is trained, inputting the sample bank card and the corresponding label information of whether the sample bank card is the abnormal card into the abnormal card identification model, outputting a training result of whether the sample bank card is the abnormal card by the abnormal card identification model, wherein the result is a probability value of whether the sample bank card is the abnormal card, calculating a corresponding loss function value according to the training result and the corresponding label information of whether the bank card is the abnormal card, and adjusting parameters of the abnormal card identification model according to the loss function value. And when the loss function value meets the requirement, completing the training of the abnormal card identification model.
After the abnormal card identification model is trained, inputting the characteristic vector of the bank card into the abnormal card identification model, and determining the probability value of the bank card as the abnormal card based on the abnormal card identification model. This probability value is one of the parameters that determines the value of the risk exposure parameter.
And determining the target similarity between the bank card and the similar transacted bank card according to the characteristic vectors corresponding to the bank card and the similar transacted bank card. According to the characteristic vectors corresponding to the bank card and the similar transacted bank card, determining each similarity of the bank card and the similar transacted bank card respectively, and then determining the target similarity according to each similarity. Here, the average value of the respective similarity degrees may be set as the target similarity degree. Preferably, the maximum similarity is selected from the similarities as the target similarity. The target similarity is one of the parameters that determines the value of the risk exposure parameter.
For example, the feature vector of the bank card of the current transaction is X c The feature vector of a certain bank card traded in the same class is X ck If the similarity between the two is
Wherein
Represents X c Is transposed, | X c ‖ 2 Represents X c L of 2 And (4) norm.
And counting the accumulated transaction number and the transaction interval time of the bank cards transacted in the same class, wherein the counting process of the transaction interval time can be that the interval time of every two adjacent transactions is respectively counted, and then the average value of the interval time of every two adjacent transactions is used as the transaction interval time. Preferably, the interval between the current transaction and the last transaction of the same type of bank card is used as the transaction interval.
According to the probability value, the target similarity, the accumulated transaction number andand determining the value of the risk exposure parameter at the transaction interval time, and identifying the abnormal card of the bank card according to the value of the risk exposure parameter. Specifically, the probability value, the target similarity, the accumulated transaction number and the transaction interval time are substituted into the following formula to determine a risk exposure parameter value;
in the formula, G t And f (X) is the probability value that the bank card is an abnormal card, alpha is a hyper-parameter for adjusting the weight, S is the target similarity, N is the accumulated transaction number, and T is the transaction interval time.
The abnormal card identification of the bank card according to the risk exposure parameter value comprises the following steps: and judging whether the risk exposure parameter value is larger than a preset risk threshold value, if so, determining that the bank card is an abnormal card, and if not, determining that the bank card is a normal card.
The preset risk threshold may be set according to a scene and a requirement, for example, the preset risk threshold is set to a smaller value such as 0.2, 0.3, or may be set to a larger value such as 0.5, 0.6. And if the determined risk exposure parameter value is larger than the preset risk threshold value, the card is considered to be an abnormal card, the current transaction is rejected, and the card number of the bank card in the current transaction is handed to a business person for verification. And if the determined risk exposure parameter value is not greater than the preset risk threshold value, the card is considered to be a normal card, and the current transaction is continued.
In order to avoid missing abnormal cards, in the embodiment of the invention, if the bank card is determined to be an abnormal card, all the transacted bank cards similar to the bank card are determined to be abnormal cards. And then, handing the card number of the determined abnormal card to a service worker for checking, if the card number is checked to be the abnormal card, adding a blacklist as a sample bank card in a training set, and iterating the training of the abnormal card identification model, thereby further ensuring the accuracy of the abnormal card identification model.
The method for identifying the abnormal card provided by the embodiment of the invention has the following specific technical effects and advantages:
(1) the abnormal card detection strategy provided by the embodiment of the invention regards the marketing campaign as a dynamic system, and constructs the dynamic risk exposure coefficient of the marketing campaign by adopting real-time data based on the transaction sequence of a card holder. Furthermore, the risk exposure coefficient is combined, marketing activities are detected in the process, the benefit of a company is guaranteed, meanwhile, the interference rate of normal transaction behaviors of consumers is reduced, and the user experience is improved.
(2) The dynamic risk exposure coefficient algorithm provided by the embodiment of the invention is not limited to only considering the number of cards of the same card Bin, and simultaneously combines the risks of the card holders themselves, the transaction behavior tendency of the card holders of the same card Bin, the accumulated transaction intensity and the like, can intelligently quantify the risks at different moments in marketing activities, can be better fused with an empirical method, and makes up the singleness of the existing rules.
(3) The embodiment of the invention innovatively provides an abnormal card detection strategy based on the dynamic risk exposure coefficient, integrates the multidimensional characteristics of a card holder, and lightens the dependence degree on expert opinions.
(4) The strategy provided by the embodiment of the invention has better expansibility, can be applied to various wind control scenes, and has stronger interpretability.
The following describes the abnormal card identification process provided by the embodiment of the present invention in detail with reference to the accompanying drawings.
Fig. 2 is a flowchart of an abnormal card identification process provided in the embodiment of the present invention, including the following steps:
s201: and classifying the bank cards in which the transaction occurs by taking the bank identification codes as dimensions.
S202: for each type of bank card, respectively determining a basic characteristic vector, a transaction characteristic vector and a transaction network characteristic vector of the bank card according to the historical transaction record of the bank card; and determining the characteristic vector of the bank card according to the basic characteristic vector, the transaction characteristic vector and the transaction network characteristic vector of the bank card.
S203: inputting the characteristic vector of the bank card into an abnormal card identification model which is trained in advance, and determining the probability value of the bank card as the abnormal card based on the abnormal card identification model.
S204: respectively determining each similarity of the bank card and the similar transacted bank card according to the characteristic vectors corresponding to the bank card and the similar transacted bank card; and selecting the maximum similarity from the similarity as the target similarity.
S205: counting the accumulated transaction stroke number and the transaction interval time of the same type of transacted bank card, substituting the probability value, the target similarity, the accumulated transaction stroke number and the transaction interval time into the following formula, and determining a risk exposure parameter value;
in the formula, G t And f (X) is the probability value that the bank card is an abnormal card, alpha is a hyper-parameter for adjusting the weight, S is the target similarity, N is the accumulated transaction number, and T is the transaction interval time.
S206: and judging whether the risk exposure parameter value is larger than a preset risk threshold value, if so, determining that the bank card is an abnormal card, and if not, determining that the bank card is a normal card.
The specific technical content is as follows:
1. and (4) identifying the risk of the bank card.
The bank card used for making an illegal profit has a certain difference from the normal bank card in the dimensions of transaction behaviors and the like. The embodiment of the invention is to compare and analyze the characteristics of the bank card of the card holder and the historical transaction condition and extract corresponding risk characteristics. And training a card holder risk detection model, namely an abnormal card identification model by adopting a supervised machine learning algorithm, and further obtaining the characteristic vector of the bank card participating in the marketing activity and predicting the probability value of the bank card as the abnormal card.
2. And (4) marketing activity dynamic risk exposure coefficient algorithm.
And recording a transaction sequence of the marketing campaign, and when a second or more bank card transactions of the same card Bin occur, starting to calculate a dynamic risk exposure coefficient of the marketing campaign, wherein the coefficient comprehensively considers the illegal profit risk of the cardholder, the feature similarity between the current cardholder and the bank card of the same card Bin, the transaction interval time, the accumulated transaction number of the same card Bin and the like to obtain the dynamic risk exposure coefficient of the current transaction.
3. And detecting the strategy in the abnormal card transaction.
And detecting abnormal card transaction events according to the constructed marketing activity dynamic risk exposure coefficient. If the coefficient exceeds a preset risk threshold value, the current transaction is rejected, and meanwhile, the subsequent transaction of the bank card suspected of having abnormal behaviors is rejected, so that effective defense is carried out. And handing the suspected abnormal card number to a service worker for verification, adding the verified card number into a blacklist to serve as a seed bank, and iteratively optimizing the abnormal card identification model algorithm. If the coefficient does not exceed the threshold, the current transaction is considered to be safe, the transaction should not be interfered, and the disturbance rate to the marketing campaign is reduced.
1. And bank card profit risk identification based on a supervised learning algorithm.
(1) Card holder feature extraction:
the abnormal card identification method provided by the embodiment of the invention is based on the basic information and transaction data of the bank card, compares and analyzes the abnormal card marked with abnormal behaviors and the normal card, and extracts the corresponding characteristic vector. In addition to basic characteristics and transaction characteristics of bank cards, considering the nature of the same Bin continuous card numbers in the scene, a heterogeneous transaction network with the card holder card number-merchant number as a node and transaction information as an edge can be constructed through a card holder transaction sequence. Thus, the cardholder feature vector dimensions may include: the specific structure of the basic feature vector of the bank card, the transaction feature vector of the card holder, the transaction network feature vector and the like is shown in fig. 3.
(2) Bank card profit risk identification based on supervised learning algorithm:
the method is characterized in that a supervised learning algorithm is constructed to identify abnormal cards, such as a Gradient Boosting Decision Tree (GBDT), and an algorithm for classifying or regressing data is achieved by adopting an addition model and continuously reducing residual errors generated in a training process. The classifier is equivalent to a mapping f:
wherein X i Is the feature vector of the ith bank card, Y i As labels for bank cards, Y i 1 denotes an abnormal card, Y i 0 denotes a normal card and N is the training set size. The feature vector X of the ith bank card i As input, label Y of bank card i And (4) as output, training an abnormal card identification model, and predicting the probability value f (X) that the bank card participating in the marketing activity is the abnormal card.
2. And (4) marketing activity dynamic risk exposure coefficient algorithm.
Regarding the marketing campaign as a dynamic system, calculating the risk exposure coefficient of the marketing campaign step by step according to the transaction behavior sequence of the cardholder, namely, when there is a second or more bank card transactions of the same card Bin, the calculation is started, and the specific algorithm of the risk exposure coefficient is as follows:
(1) calculating the probability value f (X) that the bank card c of the current transaction is an abnormal card c )。
(2) Calculating the bank card c of the current transaction and the bank cards c of all the previous same cards Bin k (K-1, …, K) is the maximum similarity coefficient S. Calculating the similarity between the bank cards according to the cosine similarity, and giving the bank cards c and c k The feature vectors are X respectively c And
calculating bank cards c and c k Has cosine similarity of
Wherein
Represents X c Is transposed, | X c ‖ 2 Represents X c L of 2 And (4) norm. In general, sim (c, c) k ) The higher the correlation degree of the two bank cards is, the more likely the abnormal cards belonging to the same organization are, and the risk of the transaction should be increased.
Further, taking the bank card c of the current transaction and the bank cards c of all the previous same cards Bin k The maximum value of the similarity coefficient of (K ═ 1, …, K) is denoted as S ═ Max k＝1,…,K sim(c,c k )。
(3) And calculating the interval time between the current transaction and the previous bank card transaction of the same card Bin, and recording the interval time as T, and the accumulated transaction number of the same card Bin as N. Intuitively, the abnormal card is more characterized by a plurality of transactions in a short time and continuously, so that the increase of T and N indicates the increase of the risk of the transactions.
In conclusion, the risk exposure coefficient G of the marketing campaign at the time of the tth transaction t Comprises the following steps:
wherein α is a hyper-parameter for adjusting the weight, and can be selected according to expert experience or CV (Cross Validation). Note that in the above formula, the risk f (x) of the current bank card, the maximum feature similarity S between the current bank card and the bank card with the same card Bin, the transaction interval time T, and the accumulated transaction number N of the same card Bin are all used as penalty factors, which jointly affect the risk exposure coefficient G. The value of the coefficient G is between 0 and 1, and the coefficient G increases along with the increase of the penalty factor, namely the risk of arbitrage of the marketing campaign increases.
3. And detecting the strategy in the event of abnormal card.
Risk exposure factor G according to marketing campaign t And detecting abnormal card in fact. Setting a risk tolerance threshold value G according to historical experience of experts or merchants 0 In fact, the detection strategy is as follows:
(1) if G is t <G 0 If the current transaction is considered to be safe, the transaction is continued without interfering with the transaction, and the disturbance rate to the marketing campaign is reduced.
(2) If G is t ≥G 0 And rejecting the current transaction and rejecting the subsequent transaction of the bank card suspected of having abnormal behavior at the same time to perform effective defense.
(3) Iterative optimization: and handing the suspected abnormal card number to a service worker for verification, adding the verified abnormal card number into a blacklist after verification, and simultaneously using the blacklist as a seed bank to iteratively optimize the algorithm.
Fig. 4 is a flowchart of an overall implementation of abnormal card identification according to the embodiment of the present invention, after a marketing campaign is developed, a bank card feature vector is extracted, and a supervised learning algorithm is constructed to obtain a bank card risk f (x); when a bank card transaction of a second or more identical cards Bin occurs, the risk exposure coefficient G begins to be calculated t (ii) a If G is t <G 0 If the current transaction is considered to be safe, continuing the transaction without interfering with the transaction, and reducing the disturbance rate to the marketing campaign, if G t ≥G 0 And rejecting the current transaction and rejecting the subsequent transaction of the bank card suspected of having abnormal behavior at the same time to perform effective defense. And handing the suspected abnormal card number to a service worker for verification, adding the card number into a blacklist after verification, and simultaneously using the card number as a seed bank to iteratively optimize the algorithm.
Fig. 5 is a schematic structural diagram of an abnormal card identification apparatus according to an embodiment of the present invention, where the apparatus includes:
the classification module 51 is used for classifying the bank cards with the bank identification codes as dimensions;
the first determining module 52 is configured to determine, for each type of bank card, a feature vector of the bank card according to a historical transaction record of the bank card, input the feature vector of the bank card into a pre-trained abnormal card identification model, and determine, based on the abnormal card identification model, a probability value that the bank card is an abnormal card;
a second determining module 53, configured to determine, according to feature vectors corresponding to the bank card and the similar transacted bank card, a target similarity between the bank card and the similar transacted bank card;
a third determining module 54, configured to count the accumulated transaction number and the transaction interval time of the bank cards of the same type that have been transacted, and determine a risk exposure parameter value according to the probability value, the target similarity, the accumulated transaction number, and the transaction interval time;
and the identification module 55 is used for performing abnormal card identification on the bank card according to the risk exposure parameter value.
The first determining module 52 is specifically configured to determine a basic feature vector, a transaction feature vector, and a transaction network feature vector of the bank card according to the historical transaction record of the bank card; and determining the characteristic vector of the bank card according to the basic characteristic vector, the transaction characteristic vector and the transaction network characteristic vector of the bank card.
The device further comprises:
and the training module 56 is configured to, for each sample bank card in the training set, input the sample feature vector of the sample bank card and the corresponding label information of whether the sample bank card is an abnormal card into the abnormal card identification model, and train the abnormal card identification model.
The second determining module 53 is specifically configured to determine, according to the feature vectors corresponding to the bank card and the similar transacted bank card, respective similarities between the bank card and the similar transacted bank card; and selecting the maximum similarity from the similarity as the target similarity.
The third determining module 54 is specifically configured to substitute the probability value, the target similarity, the accumulated transaction number and the transaction interval time into the following formula to determine a risk exposure parameter value;
in the formula, G t And f (X) is a probability value that the bank card is an abnormal card, alpha is a hyperparameter for adjusting the weight, S is the target similarity, N is the accumulated transaction number, and T is the transaction interval time.
The identification module 55 is specifically configured to determine whether the risk exposure parameter value is greater than a preset risk threshold value, determine that the bank card is an abnormal card if the risk exposure parameter value is greater than the preset risk threshold value, and determine that the bank card is a normal card if the risk exposure parameter value is not greater than the preset risk threshold value.
The identification module 55 is further configured to determine all transacted bank cards similar to the bank card as abnormal cards if the bank card is determined as an abnormal card.
An embodiment of the present invention further provides an electronic device, as shown in fig. 6, including: the system comprises a processor 301, a communication interface 302, a memory 303 and a communication bus 304, wherein the processor 301, the communication interface 302 and the memory 303 complete mutual communication through the communication bus 304;
the memory 303 has stored therein a computer program which, when executed by the processor 301, causes the processor 301 to perform the steps of:
classifying the bank cards in which the transaction occurs by taking the bank identification codes as dimensions;
for each type of bank card, determining a feature vector of the bank card according to a historical transaction record of the bank card, inputting the feature vector of the bank card into a pre-trained abnormal card identification model, and determining a probability value of the bank card as an abnormal card based on the abnormal card identification model;
determining the target similarity of the bank card and the similar transacted bank card according to the characteristic vectors corresponding to the bank card and the similar transacted bank card;
counting the accumulated transaction number and the transaction interval time of the same type of transacted bank card, and determining a risk exposure parameter value according to the probability value, the target similarity, the accumulated transaction number and the transaction interval time;
and performing abnormal card identification on the bank card according to the risk exposure parameter value.
Based on the same inventive concept, the embodiment of the present invention further provides an electronic device, and since the principle of solving the problem of the electronic device is similar to that of the abnormal card identification method, the implementation of the electronic device may refer to the implementation of the method, and the repeated parts are not described again.
The electronic device provided by the embodiment of the invention can be a desktop computer, a portable computer, a smart phone, a tablet computer, a Personal Digital Assistant (PDA), a network side device and the like.
The communication bus mentioned in the electronic device may be a Peripheral Component Interconnect (PCI) bus, an Extended Industry Standard Architecture (EISA) bus, or the like. The communication bus may be divided into an address bus, a data bus, a control bus, etc. For ease of illustration, only one thick line is shown, but this does not mean that there is only one bus or one type of bus.
The communication interface 302 is used for communication between the above-described electronic apparatus and other apparatuses.
The Memory may include a Random Access Memory (RAM) or a Non-Volatile Memory (NVM), such as at least one disk Memory. Alternatively, the memory may be at least one memory device located remotely from the processor.
The Processor may be a general-purpose Processor, including a central processing unit, a Network Processor (NP), and the like; but may also be a Digital Signal Processor (DSP), an application specific integrated circuit, a field programmable gate array or other programmable logic device, discrete gate or transistor logic, discrete hardware components, or the like.
An embodiment of the present invention further provides a computer storage readable storage medium, in which a computer program executable by an electronic device is stored, and when the program runs on the electronic device, the electronic device is caused to execute the following steps:
classifying the bank cards in which the transaction occurs by taking the bank identification code as a dimension;
for each type of bank card, determining a feature vector of the bank card according to a historical transaction record of the bank card, inputting the feature vector of the bank card into a pre-trained abnormal card identification model, and determining a probability value of the bank card as an abnormal card based on the abnormal card identification model;
determining the target similarity of the bank card and the similar transacted bank card according to the characteristic vectors corresponding to the bank card and the similar transacted bank card;
counting the accumulated transaction number and the transaction interval time of the bank cards transacted in the same category, and determining a risk exposure parameter value according to the probability value, the target similarity, the accumulated transaction number and the transaction interval time;
and performing abnormal card identification on the bank card according to the risk exposure parameter value.
Based on the same inventive concept, embodiments of the present invention further provide a computer-readable storage medium, and since a principle of solving a problem when a processor executes a computer program stored in the computer-readable storage medium is similar to that of the abnormal card identification method, implementation of the computer program stored in the computer-readable storage medium by the processor may refer to implementation of the method, and repeated details are not repeated.
The computer readable storage medium may be any available medium or data storage device that can be accessed by a processor in an electronic device, including but not limited to magnetic memory such as floppy disks, hard disks, magnetic tape, magneto-optical disks (MOs), etc., optical memory such as CDs, DVDs, BDs, HVDs, etc., and semiconductor memory such as ROMs, EPROMs, EEPROMs, non-volatile memory (NAND FLASH), Solid State Disks (SSDs), etc.
The present invention has been described with reference to flowchart illustrations and/or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the invention. It will be understood that each flow and/or block of the flow diagrams and/or block diagrams, and combinations of flows and/or blocks in the flow diagrams and/or block diagrams, can be implemented by computer program instructions. These computer program instructions may be provided to a processor of a general purpose computer, special purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, create means for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.
These computer program instructions may also be stored in a computer-readable memory that can direct a computer or other programmable data processing apparatus to function in a particular manner, such that the instructions stored in the computer-readable memory produce an article of manufacture including instruction means which implement the function specified in the flowchart flow or flows and/or block diagram block or blocks.
These computer program instructions may also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer implemented process such that the instructions which execute on the computer or other programmable apparatus provide steps for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.
While preferred embodiments of the present invention have been described, additional variations and modifications in those embodiments may occur to those skilled in the art once they learn of the basic inventive concepts. Therefore, it is intended that the appended claims be interpreted as including preferred embodiments and all such alterations and modifications as fall within the scope of the invention.
It will be apparent to those skilled in the art that various changes and modifications may be made in the present invention without departing from the spirit and scope of the invention. Thus, if such modifications and variations of the present invention fall within the scope of the claims of the present invention and their equivalents, the present invention is also intended to include such modifications and variations.
Claims (16)
1. An abnormal card identification method, characterized in that the method comprises:
classifying the bank cards in which the transaction occurs by taking the bank identification codes as dimensions;
for each type of bank card, determining a feature vector of the bank card according to a historical transaction record of the bank card, inputting the feature vector of the bank card into a pre-trained abnormal card identification model, and determining a probability value of the bank card as an abnormal card based on the abnormal card identification model;
determining the target similarity of the bank card and the similar transacted bank card according to the characteristic vectors corresponding to the bank card and the similar transacted bank card;
counting the accumulated transaction number and the transaction interval time of the same type of transacted bank card, and determining a risk exposure parameter value according to the probability value, the target similarity, the accumulated transaction number and the transaction interval time;
and performing abnormal card identification on the bank card according to the risk exposure parameter value.
2. The method of claim 1, wherein determining the feature vector of the bank card from the historical transaction record of the bank card comprises:
respectively determining a basic characteristic vector, a transaction characteristic vector and a transaction network characteristic vector of the bank card according to the historical transaction record of the bank card;
and determining the characteristic vector of the bank card according to the basic characteristic vector, the transaction characteristic vector and the transaction network characteristic vector of the bank card.
3. The method of claim 1, wherein the training process of the abnormal card recognition model comprises:
and aiming at each sample bank card in the training set, inputting the sample characteristic vector of the sample bank card and the corresponding marking information of whether the sample bank card is an abnormal card into the abnormal card identification model, and training the abnormal card identification model.
4. The method of claim 1, wherein determining the target similarity between the bank card and the like transacted bank card according to the feature vectors corresponding to the bank card and the like transacted bank card comprises:
respectively determining each similarity of the bank card and the similar transacted bank card according to the characteristic vectors corresponding to the bank card and the similar transacted bank card;
and selecting the maximum similarity from the similarity as the target similarity.
5. The method of claim 1, wherein the determining a risk exposure parameter value based on the probability value, the target similarity, the cumulative number of transactions, and the transaction interval time comprises:
substituting the probability value, the target similarity, the accumulated transaction number and the transaction interval time into the following formula to determine a risk exposure parameter value;
in the formula, G t And f (X) is the probability value that the bank card is an abnormal card, alpha is a hyper-parameter for adjusting the weight, S is the target similarity, N is the accumulated transaction number, and T is the transaction interval time.
6. The method of claim 1, wherein said performing an exception card identification of the bank card based on the risk exposure parameter value comprises:
and judging whether the risk exposure parameter value is larger than a preset risk threshold value, if so, determining that the bank card is an abnormal card, and if not, determining that the bank card is a normal card.
7. The method of claim 1, wherein the method further comprises:
and if the bank card is determined to be an abnormal card, determining all the transacted bank cards similar to the bank card as abnormal cards.
8. An abnormal card recognition apparatus, comprising:
the classification module is used for classifying the bank cards in which the transaction occurs by taking the bank identification codes as dimensions;
the first determination module is used for determining a feature vector of each type of bank card according to the historical transaction record of the bank card, inputting the feature vector of the bank card into a pre-trained abnormal card identification model, and determining the probability value of the bank card as the abnormal card based on the abnormal card identification model;
the second determining module is used for determining the target similarity of the bank card and the similar transacted bank card according to the characteristic vectors corresponding to the bank card and the similar transacted bank card;
the third determining module is used for counting the accumulated transaction stroke number and the transaction interval time of the similar transacted bank cards and determining a risk exposure parameter value according to the probability value, the target similarity, the accumulated transaction stroke number and the transaction interval time;
and the identification module is used for carrying out abnormal card identification on the bank card according to the risk exposure parameter value.
9. The apparatus according to claim 8, wherein the first determining module is specifically configured to determine a basic feature vector, a transaction feature vector, and a transaction network feature vector of the bank card according to a historical transaction record of the bank card; and determining the characteristic vector of the bank card according to the basic characteristic vector, the transaction characteristic vector and the transaction network characteristic vector of the bank card.
10. The apparatus of claim 8, wherein the apparatus further comprises:
and the training module is used for inputting the sample characteristic vectors of the sample bank cards and the corresponding marking information of whether the sample bank cards are abnormal cards or not into the abnormal card identification model aiming at each sample bank card in the training set, and training the abnormal card identification model.
11. The apparatus according to claim 8, wherein the second determining module is specifically configured to determine, according to the feature vectors corresponding to the bank card and the similar transacted bank cards, respective similarities between the bank card and the similar transacted bank cards respectively; and selecting the maximum similarity from the similarity as the target similarity.
12. Such as rightThe apparatus of claim 8, wherein the third determining module is specifically configured to substitute the probability value, the target similarity, the cumulative transaction number, and the transaction interval time into the following formula to determine a risk exposure parameter value;
in the formula, G t And f (X) is the probability value that the bank card is an abnormal card, alpha is a hyper-parameter for adjusting the weight, S is the target similarity, N is the accumulated transaction number, and T is the transaction interval time.
13. The apparatus of claim 8, wherein the identification module is specifically configured to determine whether the risk exposure parameter value is greater than a preset risk threshold, determine that the bank card is an abnormal card if yes, and determine that the bank card is a normal card if not.
14. The apparatus of claim 8, wherein the identification module is further configured to determine all transacted bank cards of the same type as the bank card as the abnormal card if the bank card is determined as the abnormal card.
15. An electronic device is characterized by comprising a processor, a communication interface, a memory and a communication bus, wherein the processor and the communication interface are used for realizing mutual communication by the memory through the communication bus;
a memory for storing a computer program;
a processor for implementing the method steps of any one of claims 1 to 7 when executing a program stored in the memory.
16. A computer-readable storage medium, characterized in that a computer program is stored in the computer-readable storage medium, which computer program, when being executed by a processor, carries out the method steps of any one of claims 1 to 7.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210527967.XA CN114881783A (en) | 2022-05-16 | 2022-05-16 | Abnormal card identification method and device, electronic equipment and storage medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210527967.XA CN114881783A (en) | 2022-05-16 | 2022-05-16 | Abnormal card identification method and device, electronic equipment and storage medium |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN114881783A true CN114881783A (en) | 2022-08-09 |
Family
ID=82674923
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202210527967.XA Pending CN114881783A (en) | 2022-05-16 | 2022-05-16 | Abnormal card identification method and device, electronic equipment and storage medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN114881783A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117195130A (en) * | 2023-09-19 | 2023-12-08 | 深圳市东陆高新实业有限公司 | Intelligent all-purpose card management system and method |
-
2022
- 2022-05-16 CN CN202210527967.XA patent/CN114881783A/en active Pending
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117195130A (en) * | 2023-09-19 | 2023-12-08 | 深圳市东陆高新实业有限公司 | Intelligent all-purpose card management system and method |
| CN117195130B (en) * | 2023-09-19 | 2024-05-10 | 深圳市东陆高新实业有限公司 | Intelligent all-purpose card management system and method |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| KR101828503B1 (en) | Apparatus and method for generating ensemble model | |
| Giudici et al. | Artificial Intelligence risk measurement | |
| US8131615B2 (en) | Incremental factorization-based smoothing of sparse multi-dimensional risk tables | |
| Dheepa et al. | Behavior based credit card fraud detection using support vector machines | |
| Suryanarayana et al. | Machine learning approaches for credit card fraud detection | |
| Fang et al. | Credit Card Fraud Detection Based on Machine Learning. | |
| Alenzi et al. | Fraud detection in credit cards using logistic regression | |
| CN114187112A (en) | Training method of account risk model and determination method of risk user group | |
| CN113689218A (en) | Risk account identification method and device, computer equipment and storage medium | |
| US11694208B2 (en) | Self learning machine learning transaction scores adjustment via normalization thereof accounting for underlying transaction score bases relating to an occurrence of fraud in a transaction | |
| Niveditha et al. | Credit card fraud detection using random forest algorithm | |
| CN113222732A (en) | Information processing method, device, equipment and storage medium | |
| Dong et al. | Real-time Fraud Detection in e-Market Using Machine Learning Algorithms. | |
| CN114881783A (en) | Abnormal card identification method and device, electronic equipment and storage medium | |
| CN111144899A (en) | Method and device for identifying false transactions and electronic equipment | |
| Mukherjee et al. | Detection of defaulters in P2P lending platforms using unsupervised learning | |
| CA3146127A1 (en) | System for detecting associated records in a record log | |
| Jain et al. | A novel approach in credit card fraud detection system using machine learning techniques | |
| CN112329862A (en) | Decision tree-based anti-money laundering method and system | |
| US11449743B1 (en) | Dimensionality reduction for statistical modeling | |
| Khang et al. | Detecting fraud transaction using ripper algorithm combines with ensemble learning model | |
| CN110570301B (en) | Risk identification method, device, equipment and medium | |
| Jamil et al. | SUSTAINABLE FRAUD DETECTION IN GREEN FINANCE EMPOWERED WITH MACHINE LEARNING APPROACH | |
| Essien | A Synergistic Approach for Enhancing Credit Card Fraud Detection using Random Forest and Naïve Bayes Models | |
| Afriyie et al. | Supervised Machine Learning Algorithm Approach to Detecting and Predicting Fraud in Credit Card Transactions |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination |