CN113793210A - Method for evaluating network loan credit, related device and computer storage medium - Google Patents
Method for evaluating network loan credit, related device and computer storage medium Download PDFInfo
- Publication number
- CN113793210A CN113793210A CN202111092909.0A CN202111092909A CN113793210A CN 113793210 A CN113793210 A CN 113793210A CN 202111092909 A CN202111092909 A CN 202111092909A CN 113793210 A CN113793210 A CN 113793210A
- Authority
- CN
- China
- Prior art keywords
- training sample
- credit
- target
- credit evaluation
- sample user
- 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
- 238000003860 storage Methods 0.000 title claims abstract description 15
- 238000000034 method Methods 0.000 title claims description 56
- 238000012549 training Methods 0.000 claims abstract description 120
- 238000011156 evaluation Methods 0.000 claims abstract description 56
- 238000013210 evaluation model Methods 0.000 claims abstract description 46
- 239000011159 matrix material Substances 0.000 claims abstract description 23
- 230000008569 process Effects 0.000 claims description 17
- 238000004590 computer program Methods 0.000 claims description 7
- 230000002776 aggregation Effects 0.000 claims description 6
- 238000004220 aggregation Methods 0.000 claims description 6
- 238000012545 processing Methods 0.000 claims description 4
- 238000009826 distribution Methods 0.000 claims description 2
- 238000000605 extraction Methods 0.000 claims description 2
- 238000010586 diagram Methods 0.000 description 8
- 230000006870 function Effects 0.000 description 5
- 238000012952 Resampling Methods 0.000 description 2
- 230000009471 action Effects 0.000 description 1
- 238000013145 classification model Methods 0.000 description 1
- 238000007418 data mining Methods 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 238000003306 harvesting Methods 0.000 description 1
- 230000010354 integration Effects 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 230000008450 motivation Effects 0.000 description 1
- 238000003062 neural network model Methods 0.000 description 1
- 230000003287 optical effect Effects 0.000 description 1
- 230000009466 transformation Effects 0.000 description 1
- 238000009827 uniform distribution Methods 0.000 description 1
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
- G06Q40/03—Credit; Loans; Processing thereof
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
Landscapes
- Engineering & Computer Science (AREA)
- Business, Economics & Management (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- Data Mining & Analysis (AREA)
- Finance (AREA)
- Accounting & Taxation (AREA)
- General Physics & Mathematics (AREA)
- Bioinformatics & Computational Biology (AREA)
- Life Sciences & Earth Sciences (AREA)
- General Engineering & Computer Science (AREA)
- Evolutionary Biology (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Artificial Intelligence (AREA)
- Evolutionary Computation (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 application provides an evaluation method of network loan credit, a related device and a computer storage medium, wherein the evaluation method of the network loan credit comprises the following steps: acquiring network loan data of a customer; inputting the network loan data into a credit evaluation model to obtain the credit evaluation grade of the customer; the credit evaluation model is obtained by training at least one classifier through a training sample set and a misclassification cost matrix; the training sample set includes: loan data and a true credit rating for at least one training sample user. Therefore, the purpose of accurately evaluating the network loan credit is achieved.
Description
Technical Field
The present application relates to the field of computer technologies, and in particular, to a method for evaluating network lending credits, a related apparatus, and a computer storage medium.
Background
Credit evaluation is the inherent demand of the development and transformation of the network lending platform, and the network lending platforms at home and abroad have the motivation and intention to establish an accurate credit evaluation method, so that the default risk and loss of lenders and platforms are reduced.
In the prior art, a data mining method is generally a main means for realizing credit evaluation of a network loan platform, and a classification model is trained to judge the credit level of a new loan application by depending on a large amount of historical data accumulated for a long time.
However, since there are more secondary loans in the online loan, the interest rate setting mode is different from the mode that the interest rate is relatively fixed in the conventional bank loan. Therefore, in the network loan, the platform not only needs to identify and reject the loan with bad credit, but also divides the loan into a plurality of credit levels according to the credit level of the borrower, and the loan interest rate difference under different credit levels is large. Resulting in inaccurate evaluation of network loan credits.
Disclosure of Invention
In view of the above, the present application provides a method, a related apparatus, and a computer storage medium for evaluating network lending credit, which are used to accurately evaluate the network lending credit.
The application provides a method for evaluating network loan credit in a first aspect, which comprises the following steps:
acquiring network loan data of a customer;
inputting the network loan data into a credit evaluation model to obtain the credit evaluation grade of the customer; the credit evaluation model is obtained by training at least one classifier through a training sample set and a misclassification cost matrix; the training sample set includes: loan data and a true credit rating for at least one training sample user.
Optionally, the method for constructing the credit evaluation model includes:
constructing a training sample set; wherein the training sample set comprises: loan data and a real credit rating of at least one training sample user;
randomly extracting a training sample user in the training sample set as a target training sample user;
aiming at each target training sample user, allocating a classifier as a target classifier;
inputting the loan data of the target training sample user into a target classifier to obtain a predicted value of the target training sample user at each credit evaluation level;
multiplying the average value of the predicted values of the target training sample users at all the credit evaluation levels by a misclassification cost matrix to obtain a prediction result;
and continuously adjusting parameters in all the target classifiers by using the errors between the prediction results and the real credit evaluation grades until the errors between the prediction results output by all the adjusted target classifiers and the real credit evaluation grades meet a preset convergence condition, and determining the combination of all the adjusted target classifiers as a credit evaluation model.
Optionally, the continuously adjusting parameters in all the target classifiers by using the error between the prediction result and the real credit evaluation level includes:
correcting the class label of the training sample user into an optimal class; wherein the best category is the category with the smallest total cost.
Optionally, the method for evaluating the network lending credit further includes:
in the process of constructing the credit evaluation model, the accuracy of the credit evaluation model is improved by utilizing a mode of guiding an aggregation algorithm to integrate learning.
A second aspect of the present application provides an apparatus for evaluating network loan credit, including:
the system comprises an acquisition unit, a processing unit and a processing unit, wherein the acquisition unit is used for acquiring the network loan data of a client;
the first input unit is used for inputting the network loan data into a credit evaluation model to obtain the credit evaluation level of the customer; the credit evaluation model is obtained by training at least one classifier through a training sample set and a misclassification cost matrix; the training sample set includes: loan data and a true credit rating for at least one training sample user.
Optionally, the credit evaluation model building unit includes:
the training sample set constructing unit is used for constructing a training sample set; wherein the training sample set comprises: loan data and a real credit rating of at least one training sample user;
the extraction unit is used for randomly extracting a training sample user from the training sample set to serve as a target training sample user;
the distribution unit is used for distributing a classifier as a target classifier for each target training sample user;
the second input unit is used for inputting the loan data of the target training sample user into a target classifier to obtain a predicted value of each credit evaluation grade of the target training sample user;
the computing unit is used for multiplying the average value of the predicted values of the target training sample users in all the credit evaluation grades by the misclassification cost matrix to obtain a prediction result;
and the model determining unit is used for continuously adjusting parameters in all the target classifiers by using the errors between the prediction results and the real credit evaluation grades, and determining the combination of all the adjusted target classifiers as a credit evaluation model when the errors between the prediction results output by all the adjusted target classifiers and the real credit evaluation grades meet the preset convergence condition.
Optionally, when the model determining unit is configured to continuously adjust parameters in all the target classifiers by using an error between the prediction result and the real credit evaluation level, the method includes:
the correcting unit is used for correcting the class label of the training sample user into the optimal class; wherein the best category is the category with the smallest total cost.
Optionally, the apparatus for evaluating network lending credit further includes:
and the unit is used for improving the accuracy of the credit evaluation model by utilizing a guide aggregation algorithm in the process of constructing the credit evaluation model.
A third aspect of the present application provides an electronic device comprising:
one or more processors;
a storage device having one or more programs stored thereon;
the one or more programs, when executed by the one or more processors, cause the one or more processors to implement the method of evaluating network lending credit of any of the first aspects.
A fourth aspect of the present application provides a computer storage medium having a computer program stored thereon, wherein the computer program, when executed by a processor, implements the method for evaluating network lending credit according to any one of the first aspects.
In view of the foregoing, the present application provides a method, a related apparatus, and a computer storage medium for evaluating network lending credits, wherein the method comprises: acquiring network loan data of a customer; inputting the network loan data into a credit evaluation model to obtain the credit evaluation grade of the customer; the credit evaluation model is obtained by training at least one classifier through a training sample set and a misclassification cost matrix; the training sample set includes: loan data and a true credit rating for at least one training sample user. Therefore, the purpose of accurately evaluating the network loan credit is achieved.
Drawings
In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the drawings needed to be used in the description of the embodiments or the prior art will be briefly introduced below, it is obvious that the drawings in the following description are only embodiments of the present application, and for those skilled in the art, other drawings can be obtained according to the provided drawings without creative efforts.
Fig. 1 is a flowchart illustrating an embodiment of a method for evaluating network credit, according to the present disclosure;
fig. 2 is a flowchart illustrating a method for constructing a credit evaluation model according to another embodiment of the present disclosure;
FIG. 3 is a diagram illustrating a method for improving the accuracy of a classifier according to another embodiment of the present application;
fig. 4 is a schematic diagram of an apparatus for evaluating network loan credit according to another embodiment of the present application;
FIG. 5 is a schematic diagram of a unit for constructing a credit evaluation model according to another embodiment of the present application;
fig. 6 is a schematic diagram of an electronic device implementing a method for evaluating network loan credit according to another embodiment of the present application.
Detailed Description
The technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the drawings in the embodiments of the present application, and it is obvious that the described embodiments are only a part of the embodiments of the present application, 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 application.
It should be noted that the terms "first", "second", and the like, referred to in this application, are only used for distinguishing different devices, modules or units, and are not used for limiting the order or interdependence of functions performed by these devices, modules or units, but the terms "include", or any other variation thereof are intended to cover a non-exclusive inclusion, so that a process, method, article, or apparatus that includes a series of elements includes not only those elements but also other elements that are not explicitly listed, or includes elements inherent to such process, method, article, or apparatus. Without further limitation, an element defined by the phrase "comprising an … …" does not exclude the presence of other identical elements in a process, method, article, or apparatus that comprises the element.
The embodiment of the application provides a method for evaluating network loan credit, as shown in fig. 1, which specifically includes the following steps:
and S101, acquiring the network loan data of the customer.
The network loan data includes information related to the loan application (such as loan amount, loan term, loan purpose, etc.), financial conditions of the borrower (such as annual income of the borrower and property of the house, etc.), and credit information of the borrower (such as credit line of the credit card account of the borrower and default conditions, etc.), which is not limited herein.
And S102, inputting the network loan data into a credit evaluation model to obtain the credit evaluation level of the client.
The credit evaluation model is obtained by training at least one classifier through a training sample set and a misclassification cost matrix; the training sample set includes: loan data and a true credit rating for at least one training sample user.
Note that, in the present application, the credit evaluation level is divided in a manner that combines the default probability and the interest rate, as shown in table 1. The credit rating of the borrower is divided into a plurality of grades, wherein the grade 1(A) represents the optimal credit, the sample is the optimal loan and the default probability PD1Minimum, interest rate I1And is also the lowest. The credit grades are gradually worsened in sequence until "n" refers to the credit level of the worst grade, the sample is bad loan, and the default probability PD isnMaximum interest rate InAnd is also largest. Therefore, generally, the worse the credit rating, the higher the probability of default and the higher the interest rate, i.e. for any i, j 1, 2.
PDi＞PDj，Ii＞Ij。
TABLE 1
Optionally, in another embodiment of the present application, an implementation manner of the method for constructing the credit evaluation model, as shown in fig. 2, includes:
s201, constructing a training sample set.
Wherein, training the sample set includes: loan data and a true credit rating for at least one training sample user.
S202, randomly extracting a training sample user in the training sample set to serve as a target training sample user.
And S203, aiming at each target training sample user, allocating a classifier as a target classifier.
It should be noted that, in the implementation of the present application, the classifier may be, but is not limited to, a neural network model, and is not limited herein.
And S204, inputting the loan data of the target training sample user into the target classifier to obtain the predicted value of the target training sample user at each credit evaluation grade.
Specifically, a training sample set is randomly sampled, and then each randomly sampled data subset S is obtainediAnd training the target classifiers, and predicting each target classifier on a training sample set to obtain a predicted value of each credit evaluation grade of the target training sample user.
S205, multiplying the average value of the predicted values of the target training sample user in all the credit evaluation levels by the misclassification cost matrix to obtain a prediction result.
Specifically, considering each training sample user x in the training sample set, calculating an average value of prediction probabilities of the training sample users x according to all categories j:
it should be noted that when a lender decides to loan a loan item with credit rating i, the lender will bear default probability, and if the loan is default, the lender will recover the principal amount 1-Lgd; if no default occurs, the lender harvests principal and interest. Therefore, when the lender borrows a loan with credit rating i, the expected profit (expectedReturn) ER is obtainediWatch capable of showingShown as follows:
ERi＝(1-PDi)·(1+Ii)+PDi·(1-Lgd)＝1+Ii-PDi·(Ii+Lgd)；
wherein Lgd represents the default loss rate. Specifically, for the lender, if the borrower pays off the loan in its full amount, the lender's income comes from paying out the interest in the amount. If the borrower violates the rules, the borrower is given a loss. Because the network loan is mostly small loan without mortgage, the lender can still recover the loan principal to a certain extent under the action of multi-term repayment. In this application, the proportion of principal lost by a lender when a borrower violates a default is measured by introducing a loss at default rate (Lgd).
In the cost sensitive method of multi-class classification, the corresponding misclassification cost matrix C ═ (C)ij) Is n-dimensional, where n is the number of classes. Any element c thereinijRepresenting the misclassification cost of a class i sample being misclassified as class j.
Based on table 1, for the multi-class structure of network lending, a corresponding misclassification cost matrix C is provided for the cost-sensitive classification problem of credit evaluation therein, see table 2. Where the elements of the diagonal positions of the matrix represent the cost of the samples being correctly classified, it is reasonable to set 0.
TABLE 2
As shown in Table 2, the misclassification matrix C can be decomposed into two triangular sub-matrices C bounded by matrix diagonals1And C2Then, there are:
C1is a lower triangular matrix containing samples representing the fact that loans with poor credit (the fact class is denoted i) are misclassified as better credit rating(prediction class is denoted by j), i>j。
C2Is an upper triangular matrix containing samples representing loans with actually better credit (actual class i) misclassified as having a worse credit rating (predicted class j), i.e., i<j。
And S206, judging whether the error between the prediction result and the real credit evaluation grade meets a preset convergence condition or not.
The preset convergence condition is set and changed by technicians, expert groups, authorized related personnel and the like, and is not limited herein.
Specifically, if it is determined that the error between the prediction result and the real credit evaluation level satisfies the preset convergence condition, step S207 is executed; if the error between the prediction result and the real credit evaluation level does not satisfy the predetermined convergence condition, step S208 is executed.
And S207, determining the combination of all the target classifiers as a credit evaluation model.
And S208, adjusting parameters in all the target classifiers by using the error between the prediction result and the real credit evaluation grade.
It should be noted that, in the specific implementation process of the present application, the model is not limited to be modified by using a preset convergence condition, and a certain maximum number of iterations may also be set to train the model, which is not limited herein.
Optionally, in another embodiment of the present application, an implementation manner of step S207 specifically includes:
and correcting the class label of the training sample user into the optimal class.
Among them, the best class is the class with the smallest total cost. Namely:
Optionally, in another embodiment of the present application, an implementation manner of the method for evaluating network lending credit further includes: in the process of constructing the credit evaluation model, the accuracy of the credit evaluation model is improved by utilizing a mode of guiding an aggregation algorithm to integrate learning.
Specifically, the accuracy of the classifier is significantly improved by using ensemble learning, and overfitting can be effectively avoided, so that the robustness of the model is improved, as shown in fig. 3. By creating various models on different random samples of the training sample set and then clustering these models to obtain the final result. This strategy works well when the underlying learning algorithm is unstable and tends to react to small changes in sample values within the input space. Therefore, due to the resampling process, duplicate entries may be typically contained in the samples, and some of the samples in the original data may be lost even if the size of the extracted samples and the original data set are the same. In particular, if we have n samples, the resampling process will take the samples with the put back evenly, where the probability that each sample is selected is 1/n per sample. Although the training subsets finally selected are different, they are not statistically completely independent and have a uniform distribution in a sense. But the existing difference causes the diversity among models in the Bagging integration process.
According to the scheme, the application provides a method for evaluating network loan credit, which comprises the following steps: acquiring network loan data of a customer; inputting the network loan data into a credit evaluation model to obtain the credit evaluation level of the client; the credit evaluation model is obtained by training at least one classifier through a training sample set and a misclassification cost matrix; the training sample set includes: loan data and a true credit rating for at least one training sample user. Therefore, the purpose of accurately evaluating the network loan credit is achieved.
Another embodiment of the present application provides an apparatus for evaluating network loan credit, as shown in fig. 4, including:
an obtaining unit 401 is configured to obtain network loan data of a customer.
The first input unit 402 is used for inputting the network loan data into the credit evaluation model to obtain the credit evaluation level of the customer.
The credit evaluation model is obtained by training at least one classifier through a training sample set and a misclassification cost matrix; the training sample set includes: loan data and a true credit rating for at least one training sample user.
For a specific working process of the unit disclosed in the above embodiment of the present application, reference may be made to the content of the corresponding method embodiment, as shown in fig. 1, which is not described herein again.
Optionally, in another embodiment of the present application, an implementation manner of the building unit of the credit evaluation model, as shown in fig. 5, includes:
a training sample set constructing unit 501, configured to construct a training sample set.
Wherein, training the sample set includes: loan data and a true credit rating for at least one training sample user.
The extracting unit 502 is configured to randomly extract a training sample user from the training sample set as a target training sample user.
An assigning unit 503, configured to assign a classifier as the target classifier for each target training sample user.
The second input unit 504 is configured to input the loan data of the target training sample user into the target classifier, so as to obtain a predicted value of each credit evaluation level of the target training sample user.
And the calculating unit 505 is configured to multiply the average value of the predicted values of the target training sample users at all the credit evaluation levels by the misclassification cost matrix to obtain a prediction result.
And the model determining unit 506 is configured to continuously adjust parameters in all the target classifiers by using errors between the prediction results and the real credit evaluation levels, and determine a combination of all the adjusted target classifiers as the credit evaluation model until the errors between the prediction results output by all the adjusted target classifiers and the real credit evaluation levels meet a preset convergence condition.
For a specific working process of the unit disclosed in the above embodiment of the present application, reference may be made to the content of the corresponding method embodiment, as shown in fig. 2, which is not described herein again.
Optionally, in another embodiment of the present application, the model determining unit 505 is configured to use an error between the prediction result and the real credit evaluation level to continuously adjust parameters in all target classifiers, and includes:
the correcting unit is used for correcting the class label of the training sample user into the optimal class; among them, the best class is the class with the smallest total cost.
For specific working processes of the units disclosed in the above embodiments of the present application, reference may be made to the contents of the corresponding method embodiments, which are not described herein again.
Optionally, in another embodiment of the present application, an implementation manner of the apparatus for evaluating network loan credit further includes:
and the ensemble learning unit is used for improving the accuracy of the credit evaluation model by utilizing a guide aggregation algorithm ensemble learning mode in the process of constructing the credit evaluation model.
For specific working processes of the units disclosed in the above embodiments of the present application, reference may be made to the contents of the corresponding method embodiments, which are not described herein again.
As can be seen from the above, the present application provides an apparatus for evaluating network loan credit, comprising: the acquisition unit 401 acquires the network loan data of the customer; the first input unit 402 inputs the network loan data into the credit evaluation model to obtain the credit evaluation level of the customer; the credit evaluation model is obtained by training at least one classifier through a training sample set and a misclassification cost matrix; the training sample set includes: loan data and a true credit rating for at least one training sample user. Therefore, the purpose of accurately evaluating the network loan credit is achieved.
Another embodiment of the present application provides an electronic device, as shown in fig. 6, including:
one or more processors 601.
A storage device 602 having one or more programs stored thereon.
The one or more programs, when executed by the one or more processors 601, cause the one or more processors 601 to implement the method of evaluating network lending credit as described in any of the embodiments above.
Another embodiment of the present application provides a computer storage medium having a computer program stored thereon, wherein the computer program, when executed by a processor, implements the method for evaluating network lending credit as described in any one of the above embodiments.
In the above embodiments disclosed in the present application, it should be understood that the disclosed apparatus and method may be implemented in other ways. The apparatus and method embodiments described above are illustrative only, as the flowcharts and block diagrams in the figures illustrate the architecture, functionality, and operation of possible implementations of apparatus, methods and computer program products according to various embodiments of the present disclosure. In this regard, each block in the flowchart or block diagrams may represent a module, segment, or portion of code, which comprises one or more executable instructions for implementing the specified logical function(s). It should also be noted that, in some alternative implementations, the functions noted in the block may occur out of the order noted in the figures. For example, two blocks shown in succession may, in fact, be executed substantially concurrently, or the blocks may sometimes be executed in the reverse order, depending upon the functionality involved. It will also be noted that each block of the block diagrams and/or flowchart illustration, and combinations of blocks in the block diagrams and/or flowchart illustration, can be implemented by special purpose hardware-based systems which perform the specified functions or acts, or combinations of special purpose hardware and computer instructions.
In addition, functional modules in the embodiments of the present disclosure may be integrated together to form an independent part, or each module may exist separately, or two or more modules may be integrated to form an independent part. The functions, if implemented in the form of software functional modules and sold or used as a stand-alone product, may be stored in a computer readable storage medium. Based on such understanding, the technical solution of the present disclosure may be embodied in the form of a software product, which is stored in a storage medium and includes several instructions for causing a computer device (which may be a personal computer, a live broadcast device, or a network device) to execute all or part of the steps of the method according to the embodiments of the present disclosure. And the aforementioned storage medium includes: a U-disk, a removable hard disk, a Read-Only Memory (ROM), a Random Access Memory (RAM), a magnetic disk or an optical disk, and other various media capable of storing program codes.
Those skilled in the art can make or use the present application. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the generic principles defined herein may be applied to other embodiments without departing from the spirit or scope of the application. Thus, the present application is not intended to be limited to the embodiments shown herein but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.
Claims (10)
1. A method for evaluating network loan credit, comprising:
acquiring network loan data of a customer;
inputting the network loan data into a credit evaluation model to obtain the credit evaluation grade of the customer; the credit evaluation model is obtained by training at least one classifier through a training sample set and a misclassification cost matrix; the training sample set includes: loan data and a true credit rating for at least one training sample user.
2. The evaluation method according to claim 1, wherein the credit evaluation model is constructed by a method comprising:
constructing a training sample set; wherein the training sample set comprises: loan data and a real credit rating of at least one training sample user;
randomly extracting a training sample user in the training sample set as a target training sample user;
aiming at each target training sample user, allocating a classifier as a target classifier;
inputting the loan data of the target training sample user into a target classifier to obtain a predicted value of the target training sample user at each credit evaluation level;
multiplying the average value of the predicted values of the target training sample users at all the credit evaluation levels by a misclassification cost matrix to obtain a prediction result;
and continuously adjusting parameters in all the target classifiers by using the errors between the prediction results and the real credit evaluation grades until the errors between the prediction results output by all the adjusted target classifiers and the real credit evaluation grades meet a preset convergence condition, and determining the combination of all the adjusted target classifiers as a credit evaluation model.
3. The evaluation method according to claim 2, wherein the continuously adjusting parameters in all the target classifiers by using the error between the prediction result and the real credit evaluation level comprises:
correcting the class label of the training sample user into an optimal class; wherein the best category is the category with the smallest total cost.
4. The evaluation method according to claim 2, further comprising:
in the process of constructing the credit evaluation model, the accuracy of the credit evaluation model is improved by utilizing a mode of guiding an aggregation algorithm to integrate learning.
5. An apparatus for evaluating network loan credit, comprising:
the system comprises an acquisition unit, a processing unit and a processing unit, wherein the acquisition unit is used for acquiring the network loan data of a client;
the first input unit is used for inputting the network loan data into a credit evaluation model to obtain the credit evaluation level of the customer; the credit evaluation model is obtained by training at least one classifier through a training sample set and a misclassification cost matrix; the training sample set includes: loan data and a true credit rating for at least one training sample user.
6. The evaluation apparatus according to claim 5, wherein the unit for constructing the credit evaluation model comprises:
the training sample set constructing unit is used for constructing a training sample set; wherein the training sample set comprises: loan data and a real credit rating of at least one training sample user;
the extraction unit is used for randomly extracting a training sample user from the training sample set to serve as a target training sample user;
the distribution unit is used for distributing a classifier as a target classifier for each target training sample user;
the second input unit is used for inputting the loan data of the target training sample user into a target classifier to obtain a predicted value of each credit evaluation grade of the target training sample user;
the computing unit is used for multiplying the average value of the predicted values of the target training sample users in all the credit evaluation grades by the misclassification cost matrix to obtain a prediction result;
and the model determining unit is used for continuously adjusting parameters in all the target classifiers by using the errors between the prediction results and the real credit evaluation grades, and determining the combination of all the adjusted target classifiers as a credit evaluation model when the errors between the prediction results output by all the adjusted target classifiers and the real credit evaluation grades meet the preset convergence condition.
7. The evaluation apparatus according to claim 6, wherein the model determination unit is configured to continuously adjust parameters in all the target classifiers by using an error between the prediction result and the real credit evaluation level, and comprises:
the correcting unit is used for correcting the class label of the training sample user into the optimal class; wherein the best category is the category with the smallest total cost.
8. The evaluation device according to claim 6, further comprising:
and the ensemble learning unit is used for improving the accuracy of the credit evaluation model by utilizing a guide aggregation algorithm ensemble learning mode in the process of constructing the credit evaluation model.
9. An electronic device, comprising:
one or more processors;
a storage device having one or more programs stored thereon;
the one or more programs, when executed by the one or more processors, cause the one or more processors to implement the method for evaluating network lending credit recited in any one of claims 1 to 4.
10. A computer storage medium having a computer program stored thereon, wherein the computer program, when executed by a processor, implements the method of evaluating network lending credit of any one of claims 1 to 4.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111092909.0A CN113793210A (en) | 2021-09-17 | 2021-09-17 | Method for evaluating network loan credit, related device and computer storage medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111092909.0A CN113793210A (en) | 2021-09-17 | 2021-09-17 | Method for evaluating network loan credit, related device and computer storage medium |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN113793210A true CN113793210A (en) | 2021-12-14 |
Family
ID=78878837
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202111092909.0A Pending CN113793210A (en) | 2021-09-17 | 2021-09-17 | Method for evaluating network loan credit, related device and computer storage medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN113793210A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115456801A (en) * | 2022-09-16 | 2022-12-09 | 北京曲速科技发展有限公司 | Artificial intelligence big data wind control system, method and storage medium for personal credit |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106779755A (en) * | 2016-12-31 | 2017-05-31 | 湖南文沥征信数据服务有限公司 | A kind of network electric business borrows or lends money methods of risk assessment and model |
| CN106897918A (en) * | 2017-02-24 | 2017-06-27 | 上海易贷网金融信息服务有限公司 | A kind of hybrid machine learning credit scoring model construction method |
| CN111932367A (en) * | 2020-08-13 | 2020-11-13 | 中国银行股份有限公司 | Pre-credit evaluation method and device |
-
2021
- 2021-09-17 CN CN202111092909.0A patent/CN113793210A/en active Pending
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106779755A (en) * | 2016-12-31 | 2017-05-31 | 湖南文沥征信数据服务有限公司 | A kind of network electric business borrows or lends money methods of risk assessment and model |
| CN106897918A (en) * | 2017-02-24 | 2017-06-27 | 上海易贷网金融信息服务有限公司 | A kind of hybrid machine learning credit scoring model construction method |
| CN111932367A (en) * | 2020-08-13 | 2020-11-13 | 中国银行股份有限公司 | Pre-credit evaluation method and device |
Non-Patent Citations (1)
| Title |
|---|
| 王浩旻: "基于代价敏感和集成学习的网络借贷信用评价方法与应用", 中国博士学位论文全文数据库基础科学辑, 15 July 2020 (2020-07-15), pages 1 - 9 * |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115456801A (en) * | 2022-09-16 | 2022-12-09 | 北京曲速科技发展有限公司 | Artificial intelligence big data wind control system, method and storage medium for personal credit |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US10572885B1 (en) | Training method, apparatus for loan fraud detection model and computer device | |
| Kumar et al. | Credit risk analysis in peer-to-peer lending system | |
| CN111260189B (en) | Risk control method, risk control device, computer system and readable storage medium | |
| CN110930218A (en) | Method and device for identifying fraudulent customer and electronic equipment | |
| CN110796485A (en) | Method and device for improving prediction precision of prediction model | |
| CN113793210A (en) | Method for evaluating network loan credit, related device and computer storage medium | |
| CN112434862B (en) | Method and device for predicting financial dilemma of marketing enterprises | |
| Ji et al. | Portfolio diversification strategy via tail‐dependence clustering and ARMA‐GARCH Vine Copula approach | |
| US20210049687A1 (en) | Systems and methods of generating resource allocation insights based on datasets | |
| US20190385079A1 (en) | Correcting bias in supervised machine learning data | |
| CN116128339A (en) | Client credit evaluation method and device, storage medium and electronic equipment | |
| CN113298642B (en) | Order detection method and device, electronic equipment and storage medium | |
| Akindaini | Machine learning applications in mortgage default prediction | |
| Lee et al. | Application of machine learning in credit risk scorecard | |
| CN113052693B (en) | Data processing method and device, electronic equipment and computer readable storage medium | |
| CN112862602B (en) | User request determining method, storage medium and electronic device | |
| Turlea | Development of Rating Models under IFRS 9 | |
| Shen et al. | Modelling the predictive performance of credit scoring | |
| CN117764692A (en) | Method for predicting credit risk default probability | |
| CN118071482A (en) | Method for constructing retail credit risk prediction model and consumer credit business Scorebetad model | |
| CN117994017A (en) | Method for constructing retail credit risk prediction model and online credit service Scoredelta model | |
| CN118333739A (en) | Method for constructing retail credit risk prediction model and retail credit business Scoremult model | |
| CN117788133A (en) | Method for constructing retail credit risk prediction model and retail credit score model | |
| CN118333737A (en) | Method for constructing retail credit risk prediction model and consumer credit business Scorebetai model | |
| CN117994016A (en) | Method for constructing retail credit risk prediction model and consumer credit business Scorebeta model |
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 |