CN116630020A - Risk assessment method and device, storage medium and electronic equipment - Google Patents
Risk assessment method and device, storage medium and electronic equipment Download PDFInfo
- Publication number
- CN116630020A CN116630020A CN202310602521.3A CN202310602521A CN116630020A CN 116630020 A CN116630020 A CN 116630020A CN 202310602521 A CN202310602521 A CN 202310602521A CN 116630020 A CN116630020 A CN 116630020A
- Authority
- CN
- China
- Prior art keywords
- transaction
- risk assessment
- features
- transaction characteristics
- scoring
- 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
- 238000012502 risk assessment Methods 0.000 title claims abstract description 88
- 238000000034 method Methods 0.000 title claims abstract description 70
- 238000003066 decision tree Methods 0.000 claims abstract description 36
- 238000012549 training Methods 0.000 claims description 27
- 238000012795 verification Methods 0.000 claims description 20
- 230000004927 fusion Effects 0.000 claims description 18
- 238000012847 principal component analysis method Methods 0.000 claims description 13
- 238000007499 fusion processing Methods 0.000 claims description 6
- 238000011156 evaluation Methods 0.000 abstract description 8
- 238000005516 engineering process Methods 0.000 abstract description 6
- 238000004590 computer program Methods 0.000 description 16
- 238000010586 diagram Methods 0.000 description 13
- 230000008569 process Effects 0.000 description 9
- 238000012545 processing Methods 0.000 description 9
- 230000006870 function Effects 0.000 description 6
- 230000006399 behavior Effects 0.000 description 4
- 230000008859 change Effects 0.000 description 3
- 238000006243 chemical reaction Methods 0.000 description 3
- 238000011161 development Methods 0.000 description 3
- 230000000694 effects Effects 0.000 description 3
- 238000000605 extraction Methods 0.000 description 3
- 230000003287 optical effect Effects 0.000 description 3
- 238000012546 transfer Methods 0.000 description 3
- 230000002159 abnormal effect Effects 0.000 description 2
- 230000009471 action Effects 0.000 description 2
- 230000005540 biological transmission Effects 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 238000012544 monitoring process Methods 0.000 description 2
- 206010000117 Abnormal behaviour Diseases 0.000 description 1
- 238000004458 analytical method Methods 0.000 description 1
- 238000004364 calculation method Methods 0.000 description 1
- 230000001419 dependent effect Effects 0.000 description 1
- 230000006866 deterioration Effects 0.000 description 1
- 230000005484 gravity Effects 0.000 description 1
- 230000006872 improvement Effects 0.000 description 1
- 238000004519 manufacturing process Methods 0.000 description 1
- 238000010606 normalization Methods 0.000 description 1
- 238000007781 pre-processing Methods 0.000 description 1
- 238000012954 risk control Methods 0.000 description 1
- 238000012216 screening Methods 0.000 description 1
- 238000000638 solvent extraction Methods 0.000 description 1
- 230000003068 static effect Effects 0.000 description 1
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
- G06Q40/03—Credit; Loans; Processing thereof
-
- 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
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N5/00—Computing arrangements using knowledge-based models
- G06N5/01—Dynamic search techniques; Heuristics; Dynamic trees; Branch-and-bound
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Business, Economics & Management (AREA)
- Data Mining & Analysis (AREA)
- Physics & Mathematics (AREA)
- Accounting & Taxation (AREA)
- Finance (AREA)
- General Engineering & Computer Science (AREA)
- Evolutionary Computation (AREA)
- Artificial Intelligence (AREA)
- Marketing (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Technology Law (AREA)
- Life Sciences & Earth Sciences (AREA)
- Strategic Management (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Bioinformatics & Computational Biology (AREA)
- General Business, Economics & Management (AREA)
- Evolutionary Biology (AREA)
- Economics (AREA)
- Development Economics (AREA)
- Computational Linguistics (AREA)
- Computing Systems (AREA)
- Mathematical Physics (AREA)
- Software Systems (AREA)
- Financial Or Insurance-Related Operations Such As Payment And Settlement (AREA)
Abstract
The application discloses a risk assessment method, a risk assessment device, a storage medium and electronic equipment. Relates to the field of financial science and technology, and the method comprises the following steps: acquiring credit card transaction characteristic data of a target account, wherein the credit card transaction characteristic data comprises first characteristic data corresponding to N transaction characteristics respectively after the target account finishes a loan transaction for a preset time period; determining weight values corresponding to the N transaction characteristics respectively; based on the first characteristic data respectively corresponding to the N transaction characteristics, a target scoring model trained based on a gradient lifting decision tree model is adopted to obtain scoring values respectively corresponding to the N transaction characteristics; and obtaining a post-credit risk assessment result of the target account according to the scoring values respectively corresponding to the N transaction characteristics and the weight values respectively corresponding to the N transaction characteristics. The application solves the problems of incomplete risk identification and low evaluation accuracy after credit card credit in the related technology.
Description
Technical Field
The application relates to the field of financial science and technology, in particular to a risk assessment method, a risk assessment device, a storage medium and electronic equipment.
Background
With the continuous development of credit card business, in order to promote benign development of credit card business, credit card post-credit risk control is more remarkable, identification and early warning of credit card post-credit risk are more prominent, monitoring and management of post-credit repayment abnormal behaviors are required to adapt to development requirements, early identification is carried out on a very small part of target data in a large amount of sample data, and risk prejudgment and effective measure intervention are facilitated. However, the risk early warning related to the method can not comprehensively identify and reduce the risk, and particularly can not effectively identify the post-credit-card feature or extract and compare the post-credit-card feature on one side in the selection of the post-credit-card data feature, so that the post-credit-card risk assessment accuracy is poor.
Aiming at the problems of incomplete risk identification and low evaluation accuracy existing in the credit card post-credit risk evaluation method in the related art, no effective solution is proposed at present.
Disclosure of Invention
The application mainly aims to provide a risk assessment method, a risk assessment device, a storage medium and electronic equipment, so as to solve the problems of incomplete risk identification and low assessment accuracy after credit card credit in the risk assessment method after credit card credit in the related technology.
In order to achieve the above object, according to one aspect of the present application, there is provided a risk assessment method. The method comprises the following steps: acquiring credit card transaction characteristic data of a target account, wherein the credit card transaction characteristic data comprises first characteristic data corresponding to N transaction characteristics respectively after the target account finishes a loan transaction for a preset time, and N is an integer greater than or equal to 1; determining weight values corresponding to the N transaction characteristics respectively; based on the first feature data corresponding to the N transaction features, a pre-trained target scoring model is adopted to obtain scoring values corresponding to the N transaction features, wherein the target scoring model is obtained by training a gradient lifting decision tree model based on historical feature data and historical feature scores corresponding to the N transaction features; and obtaining a post-credit risk assessment result of the target account according to the scoring values respectively corresponding to the N transaction characteristics and the weight values respectively corresponding to the N transaction characteristics.
In order to achieve the above object, according to another aspect of the present application, there is provided a risk assessment apparatus. The device comprises: the credit card transaction characteristic data comprises first characteristic data corresponding to N transaction characteristics respectively after the target account finishes a loan transaction for a preset time, wherein N is an integer greater than or equal to 1; the determining module is used for determining weight values corresponding to the N transaction characteristics respectively; the feature scoring module is used for obtaining scoring values corresponding to the N transaction features respectively by adopting a pre-trained target scoring model based on the first feature data corresponding to the N transaction features respectively, wherein the target scoring model is obtained by training by adopting a gradient lifting decision tree model based on historical feature data and historical feature scores corresponding to the N transaction features respectively; and the risk assessment module is used for obtaining a post-credit risk assessment result of the target account according to the scoring values respectively corresponding to the N transaction characteristics and the weight values respectively corresponding to the N transaction characteristics.
In order to achieve the above object, according to another aspect of the present application, there is also provided a nonvolatile storage medium storing a plurality of instructions adapted to be loaded by a processor and to perform any one of the above risk assessment methods.
In order to achieve the above object, according to another aspect of the present application, there is also provided an electronic device including one or more processors and a memory for storing one or more programs, wherein the one or more programs, when executed by the one or more processors, cause the one or more processors to implement any one of the risk assessment methods described above.
According to the application, the following steps are adopted: acquiring credit card transaction characteristic data of a target account, wherein the credit card transaction characteristic data comprises first characteristic data corresponding to N transaction characteristics respectively after the target account finishes a loan transaction for a preset time, and N is an integer greater than or equal to 1; determining weight values corresponding to the N transaction characteristics respectively; based on the first feature data corresponding to the N transaction features, a pre-trained target scoring model is adopted to obtain scoring values corresponding to the N transaction features, wherein the target scoring model is obtained by training a gradient lifting decision tree model based on historical feature data and historical feature scores corresponding to the N transaction features; according to the scoring values respectively corresponding to the N transaction characteristics and the weighting values respectively corresponding to the N transaction characteristics, the risk assessment result after credit of the target account is obtained, and the purposes that the influence degree of different transaction characteristics on the risk after credit is distinguished by giving different weighting values to different transaction characteristics are achieved, so that the obtained risk assessment result after credit is more accurate and reliable are achieved, and the problems that the risk identification after credit is incomplete and the assessment accuracy is low in a credit card risk assessment method in the related art are solved. Thereby achieving the effect of improving the accuracy and comprehensiveness of risk assessment after credit card credit of the user.
Drawings
The accompanying drawings, which are included to provide a further understanding of the application and are incorporated in and constitute a part of this specification, illustrate embodiments of the application and together with the description serve to explain the application. In the drawings:
FIG. 1 is a flow chart of a risk assessment method provided in accordance with an embodiment of the present application; and
FIG. 2 is an alternative schematic diagram of input sample set partitioning based on a time sliding window in accordance with an embodiment of the present application;
FIG. 3 is a flow chart of an alternative risk assessment method according to an embodiment of the present application;
FIG. 4 is a schematic diagram of an alternative transaction characteristic scoring outcome in accordance with an embodiment of the present application;
FIG. 5 is a schematic diagram of another alternative online risk assessment apparatus according to an embodiment of the present application;
FIG. 6 is a schematic diagram of a risk assessment apparatus according to an embodiment of the present application;
fig. 7 is a schematic diagram of an electronic device according to an embodiment of the present application.
Detailed Description
It should be noted that, without conflict, the embodiments of the present application and features of the embodiments may be combined with each other. The application will be described in detail below with reference to the drawings in connection with embodiments.
In order that those skilled in the art will better understand the present application, a technical solution in the embodiments of the present application will be clearly and completely described below with reference to the accompanying drawings in which it is apparent that the described embodiments are only some embodiments of the present application, not all embodiments. All other embodiments, which can be made by those skilled in the art based on the embodiments of the present application without making any inventive effort, shall fall within the scope of the present application.
It should be noted that the terms "first," "second," and the like in the description and the claims of the present application and the above figures are used for distinguishing between similar objects and not necessarily for describing a particular sequential or chronological order. It is to be understood that the data so used may be interchanged where appropriate in order to describe the embodiments of the application herein. Furthermore, the terms "comprises," "comprising," and "having," and any variations thereof, are intended to cover a non-exclusive inclusion, such that a process, method, system, article, or apparatus that comprises a list of steps or elements is not necessarily limited to those steps or elements expressly listed but may include other steps or elements not expressly listed or inherent to such process, method, article, or apparatus.
It should be noted that, the related information (including, but not limited to, user equipment information, user personal information, etc.) and data (including, but not limited to, data for presentation, analyzed data, etc.) related to the present application are information and data authorized by the user or sufficiently authorized by each party. For example, an interface is provided between the system and the relevant user or institution, before acquiring the relevant information, the system needs to send an acquisition request to the user or institution through the interface, and acquire the relevant information after receiving the consent information fed back by the user or institution.
The present application will be described with reference to preferred implementation steps, and fig. 1 is a flowchart of a risk assessment method according to an embodiment of the present application, as shown in fig. 1, and the method includes the following steps:
step S101, credit card transaction characteristic data of a target account are obtained, wherein the credit card transaction characteristic data comprise first characteristic data corresponding to N transaction characteristics respectively after the target account finishes a loan transaction for a preset time, and N is an integer greater than or equal to 1.
Optionally, the credit card transaction characteristic data of the target account is obtained by: acquiring credit card transaction related data of a target account, wherein the credit card transaction related data can be acquired in real time or acquired in advance; and carrying out feature extraction processing on the credit card transaction related data, wherein the obtained first feature data corresponding to the N transaction features are used as credit card transaction feature data of the target account.
Optionally, the N transaction characteristics are transaction characteristics that are screened from the K transaction characteristics and have a greater influence on a risk assessment result after credit card lending of the account.
In an alternative embodiment, before the obtaining the credit card transaction characteristic data of the target account, the method further includes: acquiring a first sample data set, wherein the first sample data set comprises historical characteristic data of K corresponding transaction characteristics after M accounts respectively finish the predetermined time of a loan transaction, and post-loan risk assessment results respectively corresponding to the M accounts, M is an integer greater than or equal to 1, and K is an integer greater than N; based on the first sample data set, a principal component analysis method is adopted to obtain correlation coefficients corresponding to the K transaction characteristics respectively, wherein the correlation coefficients are used for indicating the degree of association between the corresponding transaction characteristics and the post-loan risk assessment result; and determining the N transaction characteristics from the K transaction characteristics according to the corresponding correlation coefficients of the K transaction characteristics.
Optionally, the predetermined time period is determined based on a time when the account completes a bandwidth transaction and a payment period, for example, a time period corresponding to a time when the account completes a bandwidth transaction and a time one month before the payment period is determined as the predetermined time period.
Optionally, the data related to credit card transactions of the M accounts, such as removing data related to credit risk, such as transaction address, transaction time, etc., from the M accounts, where the obtained account asset status, history default times, bill period numbers, unrendered period numbers, debt conditions, personal transfer transaction details (transfer transaction), personal consumption transactions (consumption transaction), application credit card records, loan transactions, etc., form a first sample data set. By extracting features from credit card transaction related data of M accounts, historical feature data of K transaction features corresponding to the M accounts is taken as a first sample data set, where the K transaction features may include, but are not limited to: account total equity, total liabilities, age, maximum individual transaction amount, three month average liabilities, three month average equity, credit card total overdraft, annual revenue, loan total, bond value held, number of violations, work industry, average month transaction number, etc.
Optionally, the credit card transaction related data of the M accounts are screened out from the credit card transaction related data corresponding to the multiple accounts, for example, only the near three calendar history data and the active transaction credit card account are selected, the credit card state abnormal account data in the multiple accounts is removed, the account data of the credit card swiping action is not carried out for the next year, and the abnormal transaction action accounts exist, so that the M accounts are obtained.
Optionally, in the case that it is determined that missing data exists in the historical feature data of the K transaction features corresponding to the M accounts, and the data missing rate is greater than a preset proportion (for example, 50%), an interpolation method is used to fill in the missing data (for example, amount data, etc.), where the interpolation method may, but is not limited to, mean interpolation, mode interpolation, etc.
Optionally, for some transaction characteristics (such as the number of violations, total credit card overdrawing, age, etc. scattered characteristics) in the historical characteristic data of the K transaction characteristics, preprocessing is performed by scaling to better classify the characteristic data, for example, tables 1 to 3 show scaling forms of the number of violations, total credit card overdrawing, age, respectively.
TABLE 1
| Number of violations | Meaning of representation | Conversion value |
| 0 times | Violating about 0 times | 0 |
| Greater than 0 times and less than or equal to 3 times | Greater than 0 times and less than or equal to 3 times | 1 |
| More than 3 times and less than 6 times | Violating more than 3 times and less than 6 times | 2 |
| Greater than or equal to 3 times | Violating about 3 times or more | 3 |
TABLE 2
| Credit card total overdraft | Meaning of representation | Conversion value |
| Less than or equal to 1000 | Total overdraft is less than or equal to 1000 | 500 |
| More than 1000 and less than or equal to 5000 | Total overdraft is more than 1000 and less than or equal to 5000 | 3000 |
| More than 5000 and less than 10000 | Total overdraft is more than 5000 and less than 10000 | 7500 |
| More than 10000 and less than or equal to 30000 | Total overdraft is more than 10000 and less than or equal to 30000 | 20000 |
| Greater than 30000 | Total overdraft is greater than 30000 | 30000 |
TABLE 3 Table 3
| Age of | Meaning of representation | Conversion value |
| Less than or equal to 20 | Age of less than or equal to 20 | 20 |
| More than 20 and less than or equal to 30 | Age of more than 20 and less than or equal to 30 | 25 |
| Greater than 30 and less than 40 | Age above 30 and below 40 | 35 |
| Greater than 40 and less than or equal to 50 | Age of greater than 40 and less than or equal to 50 | 45 |
| Greater than 50 | Age above 50 | 55 |
It will be appreciated that there are a plurality of transaction characteristics extracted based on the credit card transaction association data, but the degree of influence of each transaction characteristic on the post-credit card risk assessment result is different, and the existence of many irrelevant transaction characteristics may cause a certain interference to the post-credit card risk assessment result, thereby making the post-credit card risk assessment result inaccurate. Based on the above, K transaction characteristics with larger influence on the credit card post-credit risk assessment result are screened out from a plurality of transaction characteristics in advance by adopting a principal component analysis method, and credit card post-credit risk assessment is carried out based on the K transaction characteristics, so that the acquired post-credit risk assessment result is more accurate and reliable.
It should be noted that, the correlation coefficient obtained based on the principal component analysis method can reflect the degree of correlation between each transaction feature and the risk assessment result after credit, and the larger the correlation coefficient is, the larger the influence degree of the risk assessment result after credit is indicated.
In an optional embodiment, the obtaining, based on the first sample data set, correlation coefficients corresponding to the K transaction features respectively by using a principal component analysis method includes: carrying out fusion processing on the K transaction characteristics to obtain L fusion characteristics, wherein L is an integer greater than or equal to 1; performing fusion processing on the feature data included in the first sample data set according to the L fusion features to obtain a second sample data set; based on the first sample data set and the second sample data set, the principal component analysis method is adopted to obtain correlation coefficients corresponding to the K transaction features and the L fusion features respectively.
Optionally, the fusing processing of the K transaction features includes: fusing X transaction characteristics in the K transaction characteristics to obtain L fusion characteristics, wherein X is an integer greater than or equal to 2, and K is less than or equal to K; 2, 3, 4, … and K of the K transaction features can be fused in sequence to obtain L fusion features.
In an optional embodiment, the determining the N transaction characteristics from the K transaction characteristics according to the correlation coefficients corresponding to the K transaction characteristics includes: and determining the N transaction characteristics from the K transaction characteristics and the L fusion characteristics according to the correlation coefficients respectively corresponding to the K transaction characteristics and the L fusion characteristics.
It will be appreciated that in some cases, the plurality of transaction feature combinations may affect the post-credit risk assessment results to a greater extent than a single transaction feature may affect the post-credit risk assessment results, e.g., the total overdraft and annual revenue combination of a credit card may affect the post-credit risk assessment results to a greater extent than the total overdraft or annual revenue combination of a credit card. Based on the above, when the transaction feature selection is performed, the combination (i.e. the fusion feature) of the transaction features is included in the screening range of the transaction features, so that the screened transaction features are more accurate, comprehensive and reliable.
Step S102, determining the weight values corresponding to the N transaction characteristics respectively.
In an optional embodiment, the determining the weight values corresponding to the N transaction characteristics includes: acquiring correlation coefficients corresponding to the N transaction characteristics respectively; summing the correlation coefficients corresponding to the N transaction characteristics respectively to obtain a first summation result; and calculating the proportion of the correlation coefficient corresponding to each of the N transaction characteristics to the first summation result to obtain the weight value corresponding to each of the N transaction characteristics.
Optionally, the values corresponding to the correlation coefficients corresponding to the N transaction features obtained by the principal component analysis method are smaller, which may result in better influence degrees of the taking method on the risk assessment result after the lending of the N transaction features. Based on the above, normalization calculation is performed on the correlation coefficients respectively corresponding to the N transaction features, namely, the specific gravity of the correlation coefficient of each transaction feature accounting for the sum of the correlation coefficients respectively corresponding to the N transaction features is calculated respectively, so that the weight values respectively corresponding to the N transaction features are obtained, and the influence degree of the N transaction features on the risk assessment result after lending is better highlighted.
Step S103, based on the first feature data corresponding to the N transaction features, a pre-trained target scoring model is adopted to obtain scoring values corresponding to the N transaction features, wherein the target scoring model is trained by adopting a gradient lifting decision tree model based on historical feature data and historical feature scores corresponding to the N transaction features.
In an optional embodiment, before the scoring values corresponding to the N transaction features are obtained by using a pre-trained target scoring model based on the first feature data corresponding to the N transaction features, the method further includes: based on a preset time sliding window, determining an input sample set from historical feature data and historical feature scores corresponding to the N transaction features respectively to obtain P input sample sets, wherein P is an integer greater than or equal to 1; presetting a first number of input sample sets in the P input sample sets as training set data, and presetting a second number of input sample sets as verification set data; sequentially inputting the preset first number of input sample sets into the gradient lifting decision tree model for training according to the time sequence corresponding to the P input sample sets, so as to obtain a trained gradient lifting decision tree model; and sequentially inputting the preset second number of input sample sets into the trained gradient lifting decision tree model according to the time sequence to verify, and taking the trained gradient lifting decision tree model as the target scoring model under the condition that verification is passed.
Optionally, the time sliding window is used to indicate the time dimension and the data amount of the word model input data. Taking the time sliding window as an example for 2 months, in the process of model training, data in the duration of 2 months is taken as an input sample set. It should be noted that, the risk assessment result after credit card lending is affected by a time factor, that is, the risk after credit generally presents a certain regularity along with the time, by using the feature data and the corresponding feature score in the predetermined period as the model input data in the above manner, the time factor is integrated in the model training process, so that the trained target score model can better highlight the time characteristic of the risk after credit.
It will be appreciated that the time dimension of the time sliding window is iterative, i.e. the previous time sliding window overlaps the next time sliding window in the time dimension, for example, for data in 2 consecutive months as a basic window, the 1 year time is divided into 11 windows, and 11 input sample sets are obtained correspondingly. The historical feature data and the historical feature scores corresponding to the N transaction features of 12 months are divided according to the data set with 2 months as a time window, and the divided input sample sets are shown in fig. 2.
Optionally, the objective function of the gradient lifting decision tree model is composed of two parts, namely, a model error, namely, a difference value between a sample true value and a predicted value, and a model structural error, namely, a regular term, which is used for limiting the complexity of the model.
Optionally, setting parameters such as the number of trees, the tree depth, the learning rate and the like corresponding to the gradient lifting decision tree model according to actual needs, for example, setting the number of trees to be 50, setting the tree depth to be 6, and setting the learning rate to be 0.30; or the number of trees is set to 80, the tree depth is 6, the learning rate is 0.30, and the method is not particularly limited.
Optionally, the model training process includes two parts of model training and model verification, the obtained P input sample sets are divided into training set data and verification set data according to a certain proportion, for example, 70% of the P input sample sets are used as training set data, 30% of the P input sample sets are used as verification set data, and the data dimension of each model input is one input sample set. Firstly, model training is carried out based on 70% serving as a training set and 30% serving as a verification set, and a trained gradient lifting decision tree model is obtained; and performing model verification through 30% of verification set data, and taking the trained gradient lifting decision tree model as a target scoring model under the condition that the trained gradient lifting decision tree model passes the verification.
Optionally, the method for verifying the trained gradient boost decision tree model may be multiple, for example, in the case that the difference between the model predicted score and the corresponding actual score is less than a preset difference threshold, determining that the trained gradient boost decision tree model passes the verification; or calculating model loss based on the model prediction score and the corresponding actual score, and determining that the trained gradient lifting decision tree model passes verification under the condition that the model loss meets the preset loss condition, wherein the specific verification method is not particularly limited.
Step S104, obtaining a post-credit risk assessment result of the target account according to the scoring values respectively corresponding to the N transaction characteristics and the weight values respectively corresponding to the N transaction characteristics.
In an optional embodiment, the obtaining the post-credit risk assessment result of the target account according to the scoring values corresponding to the N transaction features and the weight values corresponding to the N transaction features includes: obtaining a comprehensive grading value according to the grading values respectively corresponding to the N transaction characteristics and the weight values respectively corresponding to the N transaction characteristics; judging whether the comprehensive grading value is larger than a preset grading threshold value or not; and under the condition that the comprehensive grading value is larger than the preset grading threshold value, determining that the risk assessment result after the lending is that the target account has expected repayment risk.
Optionally, the scoring values corresponding to the N transaction features can be obtained through the target scoring model, the scoring values corresponding to the N transaction features are weighted according to the weighting values corresponding to the N transaction features obtained through the principal component analysis method, the comprehensive scoring value is obtained, the comprehensive scoring value is compared with a corresponding preset scoring threshold, and if the comprehensive scoring value is greater than the preset scoring threshold, the expected repayment risk of the target account is determined.
It can be appreciated that the weight values corresponding to the N transaction features can reflect the degree of influence of each transaction feature on the risk after lending. By the method, when the comprehensive score value is calculated, the influence degree of each transaction characteristic on the risk after the credit is considered, and the obtained risk assessment result after the credit is more accurate, comprehensive and reliable.
Through the steps S101 to S104, the purpose of distinguishing the influence degree of different transaction characteristics on the risk after credit by giving different weight values to different transaction characteristics so as to make the obtained risk assessment result after credit more accurate and reliable can be achieved, and the problems of incomplete risk identification and low assessment accuracy existing in the credit card risk assessment method in the related art are solved. Thereby achieving the effect of improving the accuracy and comprehensiveness of risk assessment after credit card credit of the user.
Based on the embodiment and the optional embodiment, the present application proposes an optional implementation manner, and fig. 3 is a flowchart of an optional risk assessment method according to an embodiment of the present application, and as shown in fig. 3, the method mainly includes a model training stage and a model online stage, and specifically includes:
and S1, extracting key transaction characteristics and determining weight values. N transaction characteristics with larger correlation with the risk after credit card lending are screened from K transaction characteristics with the risk after credit card lending, and the method specifically comprises the following substeps:
step S11, obtaining credit card transaction related data of M accounts, and removing data, such as transaction addresses, transaction time and the like, which are irrelevant to credit card credit risks by performing feature extraction processing on the credit card transaction related data of M accounts to obtain K transaction features and historical feature data corresponding to the K transaction features respectively, wherein the K transaction features can include, but are not limited to: account total equity, total liabilities, age, maximum individual transaction amount, three month average liabilities, three month average equity, credit card total overdraft, annual revenue, loan total, bond value held, number of violations, work industry, average month transaction number, etc.;
Step S12, historical feature data corresponding to K transaction features respectively corresponding to M accounts and post-credit risk assessment results corresponding to M accounts are used as a first sample data set, the post-credit risk assessment results corresponding to M accounts are used as dependent variables based on the first sample data set, the K transaction features are used as independent variables, a component analysis method is adopted to determine correlation coefficients corresponding to the K transaction features respectively, wherein the correlation coefficients are used for indicating the correlation degree of the transaction features and the post-credit risk of the credit card (namely, the influence degree of the transaction features on the post-credit risk of the credit card);
and S13, determining N transaction characteristics with the largest correlation coefficient from the K transaction characteristics, acquiring the correlation coefficients corresponding to the N transaction characteristics respectively, and recalculating the weight values corresponding to the N transaction characteristics respectively according to the correlation coefficients corresponding to the N transaction characteristics, namely calculating the proportion of the correlation coefficient of each transaction characteristic to the sum of the correlation coefficients corresponding to the N transaction characteristics respectively.
And S2, determining a feature scoring model. Based on a preset time sliding window, determining an input sample set from historical feature data and historical feature scores corresponding to N transaction features respectively to obtain P input sample sets, wherein P is an integer greater than or equal to 1; presetting a first number of input sample sets in the P input sample sets as training set data, and presetting a second number of input sample sets as verification set data; sequentially inputting a preset first number of input sample sets into the gradient lifting decision tree model for training according to the time sequence corresponding to the P input sample sets, so as to obtain a trained gradient lifting decision tree model; sequentially inputting a preset second number of input sample sets into the trained gradient lifting decision tree model according to the time sequence for verification, and taking the trained gradient lifting decision tree model as a target scoring model under the condition that verification is passed.
And S3, evaluating risk after online lending. The method specifically comprises the following substeps:
step S31, acquiring credit card transaction related data of a target account online, wherein the credit card transaction related data can be acquired in real time or acquired in advance; feature extraction processing is carried out on the credit card transaction associated data, and the obtained first feature data corresponding to N transaction features respectively serve as credit card transaction feature data of a target account;
step S32, based on the first feature data corresponding to the N transaction features, a pre-trained target scoring model is adopted to obtain scoring values corresponding to the N transaction features, the corresponding scoring results are shown in fig. 4, wherein the target scoring model is obtained by training through a gradient lifting decision tree model based on the historical feature data and the historical feature scores corresponding to the N transaction features;
step S33, obtaining a comprehensive grading value according to the grading values respectively corresponding to the N transaction characteristics and the weight values respectively corresponding to the N transaction characteristics; judging whether the comprehensive scoring value is larger than a preset scoring threshold value or not; and under the condition that the comprehensive grading value is larger than the preset grading threshold value, determining that the risk assessment result after the lending is that the target account has expected repayment risk, and generating a risk report and an early warning list based on the risk assessment result after the lending.
Based on the embodiment and the optional embodiment, the present application proposes an optional implementation, and fig. 5 is a schematic diagram of another optional online risk assessment device according to the embodiment of the present application, as shown in fig. 5, where the device mainly includes: the system comprises a real-time transaction behavior data acquisition module, a post-loan risk assessment module, a risk report intelligent generation module and a risk client tracking module, and is characterized in that:
the real-time transaction behavior data acquisition phase module is used for acquiring personal transaction behavior data of an account in real time, and extracting feature data corresponding to N transaction features such as personal transfer transaction, personal consumption transaction, credit card application, lending transaction data and the like;
the post-credit risk assessment module is used for leading feature data corresponding to the N transaction features to enter and exit a pre-trained target scoring model to obtain scoring values corresponding to the N transaction features respectively, and obtaining comprehensive scoring values of the account according to the scoring values corresponding to the N transaction features respectively and a pre-determined weight value, wherein the weight value is used for indicating the association degree of the transaction features and the post-credit risk of the credit card;
the risk report intelligent generation module is used for comparing the comprehensive grading value with a preset grading threshold value, and if the comprehensive grading value is larger than the preset grading threshold value, determining that the expected repayment risk exists and generating a risk report; if the comprehensive grading value is smaller than or equal to a preset grading threshold value, determining that the expected repayment risk does not exist, and continuing to detect;
And the risk client tracking module is used for extracting the account with the expected risk, generating a corresponding risk client list and sending a risk prompt to the account with the expected risk.
In the embodiment of the application, the transaction characteristics with larger risk correlation with the credit card after being credited are screened through the principal component analysis method, daily monitoring is carried out on the transaction characteristics, a change threshold is set, customers with higher credit card overdue scores are estimated by the model in a specific time before each period of repayment (such as the first three days), customers are reminded of timely repayment through channels such as short messages, the change of customer behaviors after reminding is tracked, and repayment is promoted by means of manual outbound and the like for the customers within 3 days of actual overdue, so that bad account risk is reduced. According to the embodiment of the application, not only can the overdue clients estimated by the model be tracked, but also the clients with overdue risks can be early warned in time and correspondingly intelligently reminded, the overdue clients which are overdue but not longer than 3 days are promoted to pay, the risk of N1 deterioration of the clients is reduced, the clients are timely reminded of using the standard card through short messages or voices, and for the clients which are still overdue after being reminded for many times, the measures such as derating or cancelling the card are needed to be timely adopted, the economic capital occupation amount is reduced, the damage is timely stopped, and the asset quality is effectively stabilized.
It should be noted that the steps illustrated in the flowcharts of the figures may be performed in a computer system such as a set of computer executable instructions, and that although a logical order is illustrated in the flowcharts, in some cases the steps illustrated or described may be performed in an order other than that illustrated herein.
The embodiment of the application also provides a risk assessment device, and the risk assessment device of the embodiment of the application can be used for executing the risk assessment method provided by the embodiment of the application. The risk assessment device provided by the embodiment of the application is described below.
Fig. 6 is a schematic diagram of a risk assessment apparatus according to an embodiment of the present application. As shown in fig. 6, the apparatus includes: an acquisition module 600, a determination module 602, a feature scoring module 604, a risk assessment module 606, wherein,
the obtaining module 600 is configured to obtain credit card transaction feature data of a target account, where the credit card transaction feature data includes first feature data corresponding to N transaction features after the target account completes a loan transaction for a predetermined period of time, where N is an integer greater than or equal to 1;
the determining module 602, coupled to the obtaining module 600, is configured to determine weight values corresponding to the N transaction characteristics, respectively;
The feature scoring module 604, coupled to the determining module 602, is configured to obtain scoring values corresponding to the N transaction features respectively by using a pre-trained target scoring model based on the first feature data corresponding to the N transaction features respectively, where the target scoring model is obtained by training using a gradient lifting decision tree model based on the historical feature data and the historical feature scores corresponding to the N transaction features respectively;
the risk assessment module 606 is connected to the determination module 602, and is configured to obtain a post-credit risk assessment result of the target account according to the scoring values corresponding to the N transaction features and the weight values corresponding to the N transaction features.
In the present application, the acquiring module 600 is configured to acquire credit card transaction feature data of a target account, where the credit card transaction feature data includes first feature data corresponding to N transaction features after the target account completes a loan transaction for a predetermined period of time, where N is an integer greater than or equal to 1; the determining module 602, coupled to the obtaining module 600, is configured to determine weight values corresponding to the N transaction characteristics, respectively; the feature scoring module 604, coupled to the determining module 602, is configured to obtain scoring values corresponding to the N transaction features respectively by using a pre-trained target scoring model based on the first feature data corresponding to the N transaction features respectively, where the target scoring model is obtained by training using a gradient lifting decision tree model based on the historical feature data and the historical feature scores corresponding to the N transaction features respectively; the risk evaluation module 606 is connected to the determining module 602, and is configured to obtain a post-credit risk evaluation result of the target account according to the scoring values corresponding to the N transaction features and the weight values corresponding to the N transaction features, so as to distinguish the influence degree of different transaction features on post-credit risk by assigning different weight values to different transaction features, so that the obtained post-credit risk evaluation result is more accurate and reliable, and the problems of incomplete post-credit risk identification and low evaluation accuracy in the post-credit risk evaluation method of credit cards in the related art are solved. Thereby achieving the effect of improving the accuracy and comprehensiveness of risk assessment after credit card credit of the user.
It should be noted that each of the above modules may be implemented by software or hardware, for example, in the latter case, it may be implemented by: the above modules may be located in the same processor; alternatively, the various modules described above may be located in different processors in any combination.
It should be noted that, the above-mentioned obtaining module 600, determining module 602, feature scoring module 604, and risk assessment module 606 correspond to steps S101 to S104 in the embodiment, and the above-mentioned modules are the same as the examples and application scenarios implemented by the corresponding steps, but are not limited to the disclosure of the above-mentioned embodiments. It should be noted that the above modules may be run in a computer terminal as part of the apparatus.
It should be noted that, the optional or preferred implementation manner of this embodiment may be referred to the related description in the embodiment, and will not be repeated herein.
The risk assessment device comprises a processor and a memory, wherein the units and the like are stored in the memory as program units, and the processor executes the program units stored in the memory to realize corresponding functions.
The processor includes a kernel, and the kernel fetches the corresponding program unit from the memory. The kernel may be provided with one or more by adjusting the kernel parameters (object of the present application).
The memory may include volatile memory, random Access Memory (RAM), and/or nonvolatile memory, such as Read Only Memory (ROM) or flash memory (flash RAM), among other forms in computer readable media, the memory including at least one memory chip.
An embodiment of the present application provides a nonvolatile storage medium having a program stored thereon, which when executed by a processor, implements the risk assessment method described above.
The embodiment of the application provides a processor, which is used for running a program, wherein the risk assessment method is executed when the program runs.
As shown in fig. 7, an embodiment of the present application provides an electronic device, where the electronic device 10 includes a processor, a memory, and a program stored on the memory and executable on the processor, and the processor implements the following steps when executing the program: acquiring credit card transaction characteristic data of a target account, wherein the credit card transaction characteristic data comprises first characteristic data corresponding to N transaction characteristics respectively after the target account finishes a loan transaction for a preset time, and N is an integer greater than or equal to 1; determining weight values corresponding to the N transaction characteristics respectively; based on the first feature data corresponding to the N transaction features, a pre-trained target scoring model is adopted to obtain scoring values corresponding to the N transaction features, wherein the target scoring model is obtained by training a gradient lifting decision tree model based on historical feature data and historical feature scores corresponding to the N transaction features; and obtaining a post-credit risk assessment result of the target account according to the scoring values respectively corresponding to the N transaction characteristics and the weight values respectively corresponding to the N transaction characteristics. The device herein may be a server, PC, PAD, cell phone, etc.
The application also provides a computer program product adapted to perform, when executed on a data processing device, a program initialized with the method steps of: acquiring credit card transaction characteristic data of a target account, wherein the credit card transaction characteristic data comprises first characteristic data corresponding to N transaction characteristics respectively after the target account finishes a loan transaction for a preset time, and N is an integer greater than or equal to 1; determining weight values corresponding to the N transaction characteristics respectively; based on the first feature data corresponding to the N transaction features, a pre-trained target scoring model is adopted to obtain scoring values corresponding to the N transaction features, wherein the target scoring model is obtained by training a gradient lifting decision tree model based on historical feature data and historical feature scores corresponding to the N transaction features; and obtaining a post-credit risk assessment result of the target account according to the scoring values respectively corresponding to the N transaction characteristics and the weight values respectively corresponding to the N transaction characteristics.
Optionally, the above computer program product is further adapted to execute a program initialized with the method steps of: acquiring a first sample data set, wherein the first sample data set comprises historical characteristic data of K corresponding transaction characteristics after M accounts respectively finish the predetermined time of a loan transaction, and post-loan risk assessment results respectively corresponding to the M accounts, M is an integer greater than or equal to 1, and K is an integer greater than N; based on the first sample data set, a principal component analysis method is adopted to obtain correlation coefficients corresponding to the K transaction characteristics respectively, wherein the correlation coefficients are used for indicating the degree of association between the corresponding transaction characteristics and the post-loan risk assessment result; and determining the N transaction characteristics from the K transaction characteristics according to the corresponding correlation coefficients of the K transaction characteristics.
Optionally, the above computer program product is further adapted to execute a program initialized with the method steps of: carrying out fusion processing on the K transaction characteristics to obtain L fusion characteristics, wherein L is an integer greater than or equal to 1; performing fusion processing on the feature data included in the first sample data set according to the L fusion features to obtain a second sample data set; based on the first sample data set and the second sample data set, the principal component analysis method is adopted to obtain correlation coefficients corresponding to the K transaction features and the L fusion features respectively.
Optionally, the above computer program product is further adapted to execute a program initialized with the method steps of: and determining the N transaction characteristics from the K transaction characteristics and the L fusion characteristics according to the correlation coefficients respectively corresponding to the K transaction characteristics and the L fusion characteristics.
Optionally, the above computer program product is further adapted to execute a program initialized with the method steps of: acquiring correlation coefficients corresponding to the N transaction characteristics respectively; summing the correlation coefficients corresponding to the N transaction characteristics respectively to obtain a first summation result; and calculating the proportion of the correlation coefficient corresponding to each of the N transaction characteristics to the first summation result to obtain the weight value corresponding to each of the N transaction characteristics.
Optionally, the above computer program product is further adapted to execute a program initialized with the method steps of: based on a preset time sliding window, determining an input sample set from historical feature data and historical feature scores corresponding to the N transaction features respectively to obtain P input sample sets, wherein P is an integer greater than or equal to 1; presetting a first number of input sample sets in the P input sample sets as training set data, and presetting a second number of input sample sets as verification set data; sequentially inputting the preset first number of input sample sets into the initial gradient lifting decision tree model for training according to the time sequence corresponding to the P input sample sets, so as to obtain a trained gradient lifting decision tree model; and sequentially inputting the preset second number of input sample sets into the trained gradient lifting decision tree model according to the time sequence to verify, and taking the trained gradient lifting decision tree model as the target scoring model under the condition that verification is passed.
Optionally, the above computer program product is further adapted to execute a program initialized with the method steps of: obtaining a comprehensive grading value according to the grading values respectively corresponding to the N transaction characteristics and the weight values respectively corresponding to the N transaction characteristics; judging whether the comprehensive grading value is larger than a preset grading threshold value or not; and under the condition that the comprehensive grading value is larger than the preset grading threshold value, determining that the risk assessment result after the lending is that the target account has expected repayment risk.
It will be appreciated by those skilled in the art that embodiments of the present application may be provided as a method, system, or computer program product. Accordingly, the present application may take the form of an entirely hardware embodiment, an entirely software embodiment or an embodiment combining software and hardware aspects. Furthermore, the present application may take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, and the like) having computer-usable program code embodied therein.
The present application is described with reference to flowchart illustrations and/or block diagrams of methods, apparatus (systems) and computer program products according to embodiments of the application. It will be understood that each flow and/or block of the flowchart illustrations and/or block diagrams, and combinations of flows and/or blocks in the flowchart illustrations 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.
In one typical configuration, a computing device includes one or more processors (CPUs), input/output interfaces, network interfaces, and memory.
The memory may include volatile memory in a computer-readable medium, random Access Memory (RAM) and/or nonvolatile memory, etc., such as Read Only Memory (ROM) or flash RAM. Memory is an example of a computer-readable medium.
Computer readable media, including both non-transitory and non-transitory, removable and non-removable media, may implement information storage by any method or technology. The information may be computer readable instructions, data structures, modules of a program, or other data. Examples of storage media for a computer include, but are not limited to, phase change memory (PRAM), static Random Access Memory (SRAM), dynamic Random Access Memory (DRAM), other types of Random Access Memory (RAM), read Only Memory (ROM), electrically Erasable Programmable Read Only Memory (EEPROM), flash memory or other memory technology, compact disc read only memory (CD-ROM), digital Versatile Discs (DVD) or other optical storage, magnetic cassettes, magnetic tape magnetic disk storage or other magnetic storage devices, or any other non-transmission medium, which can be used to store information that can be accessed by a computing device. Computer-readable media, as defined herein, does not include transitory computer-readable media (transmission media), such as modulated data signals and carrier waves.
It should also be noted that the terms "comprises," "comprising," or any other variation thereof, are intended to cover a non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements does not include only those elements but may include other elements not expressly listed or inherent to such process, method, article, or apparatus. Without further limitation, an element defined by the phrase "comprising one … …" does not exclude the presence of other like elements in a process, method, article or apparatus that comprises an element.
It will be appreciated by those skilled in the art that embodiments of the present application may be provided as a method, system, or computer program product. Accordingly, the present application may take the form of an entirely hardware embodiment, an entirely software embodiment or an embodiment combining software and hardware aspects. Furthermore, the present application may take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, and the like) having computer-usable program code embodied therein.
The foregoing is merely exemplary of the present application and is not intended to limit the present application. Various modifications and variations of the present application will be apparent to those skilled in the art. Any modification, equivalent replacement, improvement, etc. which come within the spirit and principles of the application are to be included in the scope of the claims of the present application.
Claims (10)
1. A risk assessment method, comprising:
acquiring credit card transaction characteristic data of a target account, wherein the credit card transaction characteristic data comprises first characteristic data corresponding to N transaction characteristics respectively after the target account finishes a loan transaction for a preset time, and N is an integer greater than or equal to 1;
Determining weight values corresponding to the N transaction characteristics respectively;
based on the first feature data corresponding to the N transaction features, a pre-trained target scoring model is adopted to obtain scoring values corresponding to the N transaction features, wherein the target scoring model is obtained by training a gradient lifting decision tree model based on historical feature data and historical feature scores corresponding to the N transaction features;
and obtaining a post-credit risk assessment result of the target account according to the scoring values respectively corresponding to the N transaction characteristics and the weight values respectively corresponding to the N transaction characteristics.
2. The method of claim 1, wherein prior to the obtaining credit card transaction characteristic data for the target account, the method further comprises:
acquiring a first sample data set, wherein the first sample data set comprises historical characteristic data of K corresponding transaction characteristics of M accounts after the predetermined time length of a loan transaction is completed respectively, and post-loan risk assessment results corresponding to the M accounts respectively, M is an integer greater than or equal to 1, and K is an integer greater than N;
based on the first sample data set, a principal component analysis method is adopted to obtain correlation coefficients corresponding to the K transaction features respectively, wherein the correlation coefficients are used for indicating the degree of association between the corresponding transaction features and the post-loan risk assessment result;
And determining the N transaction characteristics from the K transaction characteristics according to the corresponding correlation coefficients of the K transaction characteristics.
3. The method according to claim 2, wherein the obtaining, based on the first sample data set, correlation coefficients corresponding to the K transaction features respectively using a principal component analysis method includes:
carrying out fusion processing on the K transaction characteristics to obtain L fusion characteristics, wherein L is an integer greater than or equal to 1;
performing fusion processing on the feature data included in the first sample data set according to the L fusion features to obtain a second sample data set;
based on the first sample data set and the second sample data set, the principal component analysis method is adopted to obtain correlation coefficients corresponding to the K transaction features and the L fusion features respectively.
4. The method of claim 3, wherein determining the N transaction characteristics from the K transaction characteristics according to the correlation coefficients respectively corresponding to the K transaction characteristics includes:
and determining the N transaction characteristics from the K transaction characteristics and the L fusion characteristics according to the correlation coefficients respectively corresponding to the K transaction characteristics and the correlation coefficients respectively corresponding to the L fusion characteristics.
5. The method of claim 2, wherein determining the weight values for each of the N transaction characteristics comprises:
acquiring correlation coefficients corresponding to the N transaction characteristics respectively;
summing the correlation coefficients corresponding to the N transaction characteristics respectively to obtain a first summation result;
and calculating the proportion of the correlation coefficient corresponding to each of the N transaction characteristics to the first summation result to obtain the weight value corresponding to each of the N transaction characteristics.
6. The method according to claim 1, wherein before the scoring values corresponding to the N transaction features are obtained by using a pre-trained target scoring model based on the first feature data corresponding to the N transaction features, the method further comprises:
based on a preset time sliding window, determining an input sample set from historical feature data and historical feature scores corresponding to the N transaction features respectively to obtain P input sample sets, wherein P is an integer greater than or equal to 1;
presetting a first number of input sample sets in the P input sample sets as training set data, and presetting a second number of input sample sets as verification set data;
Sequentially inputting the preset first number of input sample sets into the gradient lifting decision tree model for training according to the time sequence corresponding to the P input sample sets, so as to obtain a trained gradient lifting decision tree model;
and sequentially inputting the preset second number of input sample sets into the trained gradient lifting decision tree model according to the time sequence for verification, and taking the trained gradient lifting decision tree model as the target scoring model under the condition that verification is passed.
7. The method according to any one of claims 1 to 6, wherein the obtaining the post-credit risk assessment result of the target account according to the scoring values respectively corresponding to the N transaction features and the weight values respectively corresponding to the N transaction features includes:
obtaining a comprehensive grading value according to the grading values respectively corresponding to the N transaction characteristics and the weight values respectively corresponding to the N transaction characteristics;
judging whether the comprehensive scoring value is larger than a preset scoring threshold value or not;
and under the condition that the comprehensive grading value is larger than the preset grading threshold value, determining that the risk assessment result after the lending is that the target account has expected repayment risk.
8. A risk assessment apparatus, comprising:
the credit card transaction characteristic data comprises first characteristic data corresponding to N transaction characteristics respectively after the target account finishes a loan transaction for a preset time, wherein N is an integer greater than or equal to 1;
the determining module is used for determining weight values corresponding to the N transaction characteristics respectively;
the feature scoring module is used for obtaining scoring values corresponding to the N transaction features respectively by adopting a pre-trained target scoring model based on first feature data corresponding to the N transaction features respectively, wherein the target scoring model is obtained by training by adopting a gradient lifting decision tree model based on historical feature data and historical feature scores corresponding to the N transaction features respectively;
and the risk assessment module is used for obtaining a post-credit risk assessment result of the target account according to the scoring values respectively corresponding to the N transaction characteristics and the weight values respectively corresponding to the N transaction characteristics.
9. A non-volatile storage medium storing a plurality of instructions adapted to be loaded by a processor and to perform the risk assessment method of any one of claims 1 to 7.
10. An electronic device comprising one or more processors and a memory for storing one or more programs, wherein the one or more programs, when executed by the one or more processors, cause the one or more processors to implement the risk assessment method of any of claims 1-7.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310602521.3A CN116630020A (en) | 2023-05-25 | 2023-05-25 | Risk assessment method and device, storage medium and electronic equipment |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310602521.3A CN116630020A (en) | 2023-05-25 | 2023-05-25 | Risk assessment method and device, storage medium and electronic equipment |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN116630020A true CN116630020A (en) | 2023-08-22 |
Family
ID=87616695
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202310602521.3A Pending CN116630020A (en) | 2023-05-25 | 2023-05-25 | Risk assessment method and device, storage medium and electronic equipment |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN116630020A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117670525A (en) * | 2023-12-22 | 2024-03-08 | 广东金融学院 | Enterprise credit assessment system based on big data analysis |
-
2023
- 2023-05-25 CN CN202310602521.3A patent/CN116630020A/en active Pending
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117670525A (en) * | 2023-12-22 | 2024-03-08 | 广东金融学院 | Enterprise credit assessment system based on big data analysis |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN111967779B (en) | Risk assessment method, device and equipment | |
| Sakai et al. | Firm age and the evolution of borrowing costs: Evidence from Japanese small firms | |
| US10755280B2 (en) | Segmented data analysis using dynamic peer groupings and automated rule implementation platform | |
| US20130290167A1 (en) | System and method for credit risk management for educational institutions | |
| Buriak et al. | Reputation risk in banking: application for Ukraine | |
| Bakhach et al. | TSFDC: A trading strategy based on forecasting directional change | |
| CN110796539A (en) | Credit investigation evaluation method and device | |
| US20140278774A1 (en) | In the market model systems and methods | |
| US20170061534A1 (en) | Assessing credit risk | |
| CN112819476A (en) | Risk identification method and device, nonvolatile storage medium and processor | |
| CN116630020A (en) | Risk assessment method and device, storage medium and electronic equipment | |
| Khatun et al. | Earnings manipulation behavior in the banking industry of Bangladesh: the strategical implication of Beneish M-score model | |
| CN116450951A (en) | Service recommendation method and device, storage medium and electronic equipment | |
| US10699335B2 (en) | Apparatus and method for total loss prediction | |
| CN116362895A (en) | Financial product recommendation method, device and storage medium | |
| CN116596657A (en) | Loan risk assessment method and device, storage medium and electronic equipment | |
| CN115953235A (en) | Risk index statistical method and device, storage medium and electronic equipment | |
| CN117252677A (en) | Credit line determination method and device, electronic equipment and storage medium | |
| Comporek | The use of operational cash flow in the estimation of accrual-based earnings management | |
| KR102420952B1 (en) | Loan expansion hypothesis testing system using artificial intelligence and method using the same | |
| US12045261B2 (en) | Method and apparatus for measuring material risk in a data set | |
| Chaudhuri et al. | Operational Risk | |
| CN116362753A (en) | Transaction risk processing method, device and storage medium | |
| Vojtková et al. | Intraday liquidity modelling using statistical methods | |
| CN116862675A (en) | Credit risk assessment method, credit risk assessment device and nonvolatile storage medium |
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 |