CN111242191A - Credit rating method and device based on multi-classifier integration - Google Patents

Credit rating method and device based on multi-classifier integration Download PDF

Info

Publication number
CN111242191A
CN111242191A CN202010009078.5A CN202010009078A CN111242191A CN 111242191 A CN111242191 A CN 111242191A CN 202010009078 A CN202010009078 A CN 202010009078A CN 111242191 A CN111242191 A CN 111242191A
Authority
CN
China
Prior art keywords
credit rating
classifier model
historical
classifier
data
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
Application number
CN202010009078.5A
Other languages
Chinese (zh)
Inventor
杨占栋
陈朝明
叶振栋
任婷婷
郭文琳
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
China Construction Bank Corp
Original Assignee
China Construction Bank Corp
CCB Finetech Co Ltd
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by China Construction Bank Corp, CCB Finetech Co Ltd filed Critical China Construction Bank Corp
Priority to CN202010009078.5A priority Critical patent/CN111242191A/en
Publication of CN111242191A publication Critical patent/CN111242191A/en
Pending legal-status Critical Current

Links

Images

Classifications

    • 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
    • G06F18/217—Validation; Performance evaluation; Active pattern learning techniques
    • G06F18/2193—Validation; Performance evaluation; Active pattern learning techniques based on specific statistical tests
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F18/00—Pattern recognition
    • G06F18/20—Analysing
    • G06F18/24—Classification techniques
    • G06F18/241—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches
    • G06F18/2415—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches based on parametric or probabilistic models, e.g. based on likelihood ratio or false acceptance rate versus a false rejection rate
    • G06F18/24155—Bayesian classification
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F18/00—Pattern recognition
    • G06F18/20—Analysing
    • G06F18/24—Classification techniques
    • G06F18/243—Classification techniques relating to the number of classes
    • G06F18/24323—Tree-organised classifiers
    • 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
    • G06Q40/06—Asset management; Financial planning or analysis

Landscapes

  • Engineering & Computer Science (AREA)
  • Theoretical Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • Data Mining & Analysis (AREA)
  • General Physics & Mathematics (AREA)
  • Business, Economics & Management (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Evolutionary Biology (AREA)
  • Evolutionary Computation (AREA)
  • Bioinformatics & Computational Biology (AREA)
  • General Engineering & Computer Science (AREA)
  • Bioinformatics & Cheminformatics (AREA)
  • Artificial Intelligence (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Development Economics (AREA)
  • Finance (AREA)
  • Probability & Statistics with Applications (AREA)
  • Accounting & Taxation (AREA)
  • Operations Research (AREA)
  • Game Theory and Decision Science (AREA)
  • Human Resources & Organizations (AREA)
  • Entrepreneurship & Innovation (AREA)
  • Economics (AREA)
  • Marketing (AREA)
  • Strategic Management (AREA)
  • Technology Law (AREA)
  • General Business, Economics & Management (AREA)
  • Financial Or Insurance-Related Operations Such As Payment And Settlement (AREA)

Abstract

The invention discloses a credit rating method and a device based on multi-classifier integration, wherein the method comprises the following steps: acquiring financial characteristic data of a bond issuer; respectively inputting the financial characteristic data into a trained naive Bayes classifier model, a gradient boosting decision tree classifier model and a random forest classifier model to obtain credit rating results output by the classifier models, wherein the naive Bayes classifier model, the gradient boosting decision tree classifier model and the random forest classifier model are obtained by training according to historical financial characteristic data and historical credit rating data; and determining the credit rating of the bond issuer according to the credit rating result output by each classifier model. The credit rating method based on multi-classifier integration improves the stability and accuracy of credit rating.

Description

Credit rating method and device based on multi-classifier integration
Technical Field
The invention relates to the technical field of credit service, in particular to a credit rating method and device based on multi-classifier integration.
Background
In recent years, debt violations have gradually assumed an increasing situation. Bond default is bound to cause economic damage and emotional attack to bond investment workers, and bond issuers who present credit risks are related to various large enterprises such as central enterprises, national enterprises and famous enterprises at present. Therefore, there is a need for a method for determining the credit risk level of a bond issuer using public information on the market, which can protect the interests of investors and avoid property damage.
Disclosure of Invention
The invention provides a credit rating method and device based on multi-classifier integration in order to solve at least one technical problem in the background art.
In order to achieve the above object, according to one aspect of the present invention, there is provided a credit rating method based on multi-classifier integration, the method including:
acquiring financial characteristic data of a bond issuer;
respectively inputting the financial characteristic data into a trained naive Bayes classifier model, a gradient boosting decision tree classifier model and a random forest classifier model to obtain credit rating results output by the classifier models, wherein the naive Bayes classifier model, the gradient boosting decision tree classifier model and the random forest classifier model are obtained by training according to historical financial characteristic data and historical credit rating data;
and determining the credit rating of the bond issuer according to the credit rating result output by each classifier model.
Optionally, the financial characteristic data includes: an operating revenue growth rate of four quarters, a net profit growth rate of four quarters after the unusual profit is deducted, an operating revenue to fixed asset ratio, and a flowing asset to flowing liability ratio.
Optionally, the method further includes: acquiring historical financial characteristic data and historical credit rating data; training the naive Bayes classifier model according to historical financial feature data, historical credit rating data and a preset naive Bayes classification algorithm, and performing credit rating on the bond issuer according to the trained naive Bayes classifier model.
Optionally, the method further includes: acquiring historical financial characteristic data and historical credit rating data; training the gradient lifting decision tree classifier model according to historical financial feature data, historical credit rating data and a preset gradient lifting decision tree algorithm, and performing credit rating on the bond issuer according to the trained gradient lifting decision tree classifier model.
Optionally, the method further includes: acquiring historical financial characteristic data and historical credit rating data; training the random forest classifier model according to historical financial characteristic data, historical credit rating data and a preset random forest algorithm, and performing credit rating on a bond issuer according to the trained random forest classifier model.
In order to achieve the above object, according to another aspect of the present invention, there is provided a credit rating apparatus based on multi-classifier integration, the apparatus including:
the financial characteristic data determining unit is used for acquiring financial characteristic data of the bond issuer;
the multi-classifier credit rating unit is used for respectively inputting the financial characteristic data into a trained naive Bayes classifier model, a gradient boosting decision tree classifier model and a random forest classifier model to obtain credit rating results output by the classifier models, wherein the naive Bayes classifier model, the gradient boosting decision tree classifier model and the random forest classifier model are obtained by training according to historical financial characteristic data and historical credit rating data;
and the credit rating determination unit of the bond issuer is used for determining the credit rating of the bond issuer according to the credit rating result output by each classifier model.
Optionally, the financial characteristic data includes: an operating revenue growth rate of four quarters, a net profit growth rate of four quarters after the unusual profit is deducted, an operating revenue to fixed asset ratio, and a flowing asset to flowing liability ratio.
Optionally, the apparatus further comprises: the naive Bayes classifier model training unit is used for acquiring historical financial characteristic data and historical credit rating data; and training the naive Bayes classifier model according to historical financial characteristic data, historical credit rating data and a preset naive Bayes classification algorithm, and performing credit rating on the bond issuer according to the trained naive Bayes classifier model.
Optionally, the apparatus further comprises: the gradient lifting decision tree classifier model training unit is used for acquiring historical financial characteristic data and historical credit rating data; and training the gradient lifting decision tree classifier model according to historical financial characteristic data, historical credit rating data and a preset gradient lifting decision tree algorithm, so as to perform credit rating on the bond issuer according to the trained gradient lifting decision tree classifier model.
Optionally, the apparatus further comprises: the random forest classifier model training unit is used for acquiring historical financial characteristic data and historical credit rating data; and training the random forest classifier model according to historical financial characteristic data, historical credit rating data and a preset random forest algorithm, and performing credit rating on the bond issuer according to the trained random forest classifier model.
To achieve the above object, according to another aspect of the present invention, there is also provided a computer device including a memory, a processor and a computer program stored on the memory and executable on the processor, the processor implementing the steps of the above credit rating method based on multi-classifier integration when executing the computer program.
In order to achieve the above object, according to another aspect of the present invention, there is also provided a computer-readable storage medium storing a computer program which, when executed in a computer processor, implements the steps in the above-described credit rating method based on multi-classifier integration.
The invention has the beneficial effects that: according to financial characteristic data of bond issuers, a plurality of classifier methods including Naive Bayes (NB), Gradient Boosting Decision Trees (GBDT) and Random Forests (RF) are adopted for pattern recognition, and a credit rating result under each classifier method is obtained. And finally, a final credit rating result is obtained through a preset rule, so that the condition that a certain classifier has larger deviation of the recognition result on data of local characteristics is avoided, and the stability and the accuracy of credit rating are improved.
Drawings
In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below, and it is obvious that the drawings in the following description are some embodiments of the present invention, and for those skilled in the art, other drawings can be obtained according to these drawings without creative efforts. In the drawings:
FIG. 1 is a flow chart of a credit rating method based on multi-classifier integration according to an embodiment of the present invention;
FIG. 2 is a block diagram of a credit rating device based on multi-classifier integration according to an embodiment of the present invention;
FIG. 3 is a schematic diagram of a computer apparatus according to an embodiment of the present invention.
Detailed Description
In order to make the technical solutions of the present invention better understood, the technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention, and it is obvious that the described embodiments are only a part of the embodiments of the present invention, and not all of the embodiments. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.
As will be appreciated by one skilled in the art, embodiments of the present invention may be provided as a method, system, or computer program product. Accordingly, the present invention may take the form of an entirely hardware embodiment, an entirely software embodiment or an embodiment combining software and hardware aspects. Furthermore, the present invention 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.
It should be noted that the terms "comprises" and "comprising," and any variations thereof, in the description and claims of the present invention and the above-described drawings, 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 embodiments and features of the embodiments may be combined with each other without conflict. The present invention will be described in detail below with reference to the embodiments with reference to the attached drawings.
Fig. 1 is a flowchart of a credit rating method based on multi-classifier integration according to an embodiment of the present invention, and as shown in fig. 1, the credit rating method based on multi-classifier integration according to the present embodiment includes steps S101 to S103.
And step S101, acquiring financial characteristic data of the bond issuer.
The bond issuer in the present invention may be a company, business, or other organization. In an alternative embodiment of the invention said financial characteristic data comprises: an operating revenue growth rate of four quarters, a net profit growth rate of four quarters after the unusual profit is deducted, an operating revenue to fixed asset ratio, and a flowing asset to flowing liability ratio.
In an alternative embodiment of the invention the financial characteristic data is determined in particular by the following steps. The financial data of a bond issuer is first acquired and preprocessed. Financial data may include profitability, leverage, capital structure, asset quality, debt ability, financing ability, extent of expansion, etc., with specific being total revenue, gross profit, net profit due, gross profit on sale, net profit on sale, etc. The preprocessing includes removing duplicated data and data with a large amount of missing data, and complementing missing values in the data. And further, performing feature extraction or calculation on the preprocessed financial data to obtain financial feature data.
The Revenue Growth rate for four quarters can be calculated by the following formula:
Figure BDA0002356455430000051
wherein, Total Revenuet-1Total revenues for the Total revenues of a business' recent quarter revenuest-5The total revenue that the business was exposed five quarters ago.
The net profit Growth rate after the four quarters minus the unusual profit loss can be calculated from the following formula:
Figure BDA0002356455430000052
wherein, NPADNGLtFor the net profit after the unusual profit is deducted from the company's last season on day t,
Figure BDA0002356455430000055
the net profit after the unusual profit was deducted, which was revealed by the enterprise in the last fourth season of day t.
The revenue versus Fixed Asset ratio Fixed Asset Turnover can be calculated by the following formula:
Figure BDA0002356455430000053
wherein, Core Revenuet-iFixed Asset for revenue that stock has been revealed in the last ith quartert-iA fixed asset that was recently disclosed in the ith quarter for an enterprise.
The flowing asset to flowing liability Ratio Current Ratio can be calculated by the following formula:
Figure BDA0002356455430000054
wherein Current AssettCurrentLiability, a floating asset disclosed by the enterprise in the last quarter of day ttThe floating liability revealed for the business in the last season of day t.
In the embodiment of the invention, the financial characteristic data of the bond issuer in the step is the current financial characteristic data of the bond issuer. The invention uses historical financial characteristic data of a plurality of bond issuers in training classifier models.
And S102, respectively inputting the financial characteristic data into a trained naive Bayes classifier model, a gradient boosting decision tree classifier model and a random forest classifier model to obtain credit rating results output by the classifier models, wherein the naive Bayes classifier model, the gradient boosting decision tree classifier model and the random forest classifier model are obtained by training according to historical financial characteristic data and historical credit rating data.
In the embodiment of the invention, the corresponding classifier models are trained according to historical financial characteristic data of a plurality of bond issuers and corresponding historical credit rating data respectively according to a naive Bayes classification algorithm, a gradient boosting decision tree algorithm and a random forest algorithm. And then carrying out credit rating on the target bond issuer based on the current financial feature data of the target bond issuer according to the trained three classifier models, and respectively obtaining credit rating results output by the classifier models. In an alternative embodiment of the invention, the historical credit rating data of the bond issuer (enterprise) may be rating data made by domestic and foreign rating agencies.
And step S103, determining the credit rating of the bond issuer according to the credit rating result output by each classifier model.
In an embodiment of the present invention, the credit rating result output by each classifier model may be a credit rating level. In the embodiment of the invention, the grades output by the naive Bayes classifier model, the gradient boosting decision tree classifier model and the random forest classifier model are I in sequence1,I2,I3. The invention firstly converts the rating grade output by the classifier model into a corresponding rating score, namely:
S1＝f1(I1)
S2＝f2(I2)
S3＝f3(I3)
S1、S2、S3and the grading scores are respectively corresponding to the grading results output by the naive Bayes classifier model, the gradient boosting decision tree classifier model and the random forest classifier model. f. of1、f2、f3、f4Is a preset relation function.
In an optional embodiment of the present invention, the average calculation may be performed on the rating scores corresponding to the rating results output by the three classifier models, that is:
Figure BDA0002356455430000061
in other optional embodiments of the present invention, the final credit rating score S may be obtained by performing weighted average calculation on the rating scores corresponding to the rating results output by the three classifier models according to a preset weight ratio.
Taking S as a final credit rating score and converting the score into a credit rating grade, namely:
I＝f4(S)
the credit rating level I is the final credit rating result of the bond issuer.
From the above description, it can be seen that the present invention adopts multiple classifier methods including Naive Bayes (NB), Gradient Boosting Decision Trees (GBDT), and Random Forests (RF) according to financial feature data of bond issuers to perform pattern recognition, and obtain credit rating results under each classifier method. And finally, a final credit rating result is obtained through a preset rule, so that the condition that a certain classifier has larger deviation of the recognition result on data of local characteristics is avoided, and the stability and the accuracy of credit rating are improved.
The specific training process of the naive Bayes classifier model, the gradient boosting decision tree classifier model and the random forest classifier model of the invention is introduced below. In the embodiment of the invention, the naive Bayes classifier model, the gradient boosting decision tree classifier model and the random forest classifier model are obtained by training according to the acquired historical financial feature data of a plurality of bond issuers and the corresponding historical credit rating data as training data. The financial characteristic data comprises: revenue growth rate for four quarters, net profit growth rate for four quarters with infrequent losses deducted, revenue to fixed asset ratio, and mobile asset to mobile liability ratio, etc. In the embodiment of the present invention, the present invention may use data of one year as a sample, for example, financial characteristic data of a corporation 2000 as a financial characteristic data sample, and a credit rating level of the corporation 2000 is credit rating data corresponding to the financial characteristic data sample. From this, we obtain a set of financial feature data samples,
Figure BDA0002356455430000071
wherein the content of the first and second substances,
Figure BDA0002356455430000072
representing j financial characteristic data of the i samples.
And a set of credit rating data samples,
Figure BDA0002356455430000073
where, c-22, represents the following 22 credit rating levels,
Figure BDA0002356455430000074
representing the credit rating level in i samples, i.e. when the ith sample credit rating is j, then
Figure BDA0002356455430000075
yiThe other data in (1) is 0.
In the embodiment of the present invention, the credit rating data is generally expressed by a credit rating scale, which may be defined as the following table 1:
Figure BDA0002356455430000076
Figure BDA0002356455430000081
TABLE 1
Further the invention may assemble the above-mentioned financial characteristic data samples,
Figure BDA0002356455430000082
and a set of credit rating data samples,
Figure BDA0002356455430000083
and c is 22, and the training data is used for training a corresponding classifier model by combining a naive Bayes classification algorithm, a gradient boosting decision tree algorithm and a random forest algorithm. The training process for the three models is explained below separately.
1. Naive Bayes (NB) classifier model
For the above-mentioned financial characteristic data sample set xiFurther processing is carried out, and the specific processing is as follows:
Figure BDA0002356455430000084
wherein, the [ alpha ], [ beta ]]In order to get the whole,
Figure BDA0002356455430000085
in order to be of a known value,
Figure BDA0002356455430000086
n is a known positive number.
And finally, the set of the processed financial characteristic data samples is as follows:
Figure BDA0002356455430000087
the Naive Bayes (NB) based decision criterion is:
Figure BDA0002356455430000088
where c denotes a particular credit rating level class, y denotes a set of all classes, P (c) denotes a prior probability that c belongs to a particular credit rating level, P (x)i' | c) denotes the sample x of c at a certain credit rating leveli' conditional probability, xi' means the ith financial characteristic data from 1 to d financial characteristic data in the data.
In order to improve the calculation efficiency, the invention simplifies the formula, performs logarithm taking, does not affect the result, changes multiplication into addition, and avoids underflow data overflow caused by multiplication, namely:
Figure BDA0002356455430000091
wherein, PcAnd P (x)i' | c) is obtained by training according to known training data, and finally, a trained naive Bayes classifier model is formed. And then inputting the financial characteristic data of the target bond issuer into the trained naive Bayes classifier model to obtain a credit rating result of the naive Bayes classifier model.
2. Gradient Boosting Decision Tree (GBDT) classifier model
X and Y are input and output variables, respectively, and Y is a continuous variable, given a training data set:
D＝{(x1,y1),(x2,y2),…,(xN,yN)}
wherein N is the number of samples of the data set.
In an input space where a training data set is located, recursively dividing each region into two sub-regions, determining an output value on each sub-region, and constructing a binary decision tree, specifically comprising the following steps:
1) selecting an optimal segmentation j and an optimal segmentation point s, and solving:
Figure BDA0002356455430000092
and traversing the variable j, scanning a segmentation point s for the fixed segmentation variable j, and selecting a pair (j, s) which enables the above formula to reach the minimum value.
2) Dividing the region by the selected pair (j, s) and determining the corresponding output value:
R1(j,s)＝{x|x(j)≤s},R2(j,s)＝{x|x(j)＞s}
Figure BDA0002356455430000093
3) and continuing to call the steps 1) and 2) for the two sub-areas until a stop condition is met.
4) Dividing an input space into M regions R1，R2,…,RMAt each sheetR is a member ofmHas a fixed output value cmThe decision tree is generated as follows:
Figure BDA0002356455430000094
the Gradient Boosting Decision Tree (Gradient Boosting Decision Tree) adopts an additive model (namely linear combination of basis functions) and a forward distribution algorithm. The lifting method using decision tree as basis function is a lifting tree, which is a common decision by iterating multiple regression trees. The model of the lifting tree is:
Figure BDA0002356455430000101
wherein T (x; theta)m) Represents a decision tree, θmM is the number of trees as a parameter of the decision tree.
In the iteration of GBDT, the gradient descent approximation method is used, i.e. the negative gradient of the penalty function is used to approximate the value of the current model, i.e.,
Figure BDA0002356455430000102
as a residual approximation of the lifting tree algorithm in the regression problem. The current loss function uses the mean square error, i.e.:
Figure BDA0002356455430000103
the negative gradient is:
Figure BDA0002356455430000104
i.e. the residual error.
The specific algorithm steps are as follows:
1) initialization f0(x)＝0
2) For M is 1,2 …, M
a) Calculating the residual error, namely:
rmi＝yi-fm-1(xi),i＝1,2,…,N
b) fitting residual rmiLearning a regression tree to obtain T (x; theta)m)
c) Updating the regression tree, namely:
fm(x)＝fm-1(x)+T(x；θm)
3) and obtaining an output final gradient boosting decision tree classifier model, namely:
Figure BDA0002356455430000105
after the gradient boosting decision tree classifier model is obtained, the financial characteristic data of the target bond issuer can be input into the classifier model, and a credit rating result output by the gradient boosting decision tree classifier model is obtained.
3. Random Forest (RF) classifier model
Firstly, performing credit rating pattern recognition on the extracted features based on a decision tree, specifically adopting a C4.5 decision tree algorithm, wherein the gain rate is as follows:
Figure BDA0002356455430000111
wherein the content of the first and second substances,
Figure BDA0002356455430000112
Figure BDA0002356455430000113
training the model through supervised data to obtain a parameter thetatreeAnd then calculating with the new characteristic data and the known parameters to obtain the final credit rating mode result, namely:
I＝f(θtree,F3)
randomly selecting samples and features, establishing N decision trees, outputting a credit rating mode result by each decision tree, and determining a final credit rating mode through a voting mechanism to obtain a random forest classifier model.
As can be seen from the above description, the present invention achieves at least the following advantageous effects:
the invention provides a novel method for pattern recognition of bond issuer credit in a classifier integrating mode based on different characteristics of a plurality of classifiers, and the core of the method utilizes different characteristics of different classifiers, a specific feature extraction mode and a mode of finally determining credit rating by the integrated classifier. Therefore, the situation that the recognition result of a certain classifier has larger deviation on the data of local characteristics is avoided, and the stability and the accuracy of recognition are improved.
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 different than presented herein.
Based on the same inventive concept, the embodiment of the present invention further provides a credit rating apparatus based on multi-classifier integration, which can be used to implement the credit rating method based on multi-classifier integration described in the above embodiments, as described in the following embodiments. Since the principle of solving the problem of the credit rating apparatus based on multi-classifier integration is similar to that of the credit rating method based on multi-classifier integration, the embodiment of the credit rating apparatus based on multi-classifier integration may refer to the embodiment of the credit rating method based on multi-classifier integration, and repeated details are omitted. As used hereinafter, the term "unit" or "module" may be a combination of software and/or hardware that implements a predetermined function. Although the means described in the embodiments below are preferably implemented in software, an implementation in hardware, or a combination of software and hardware is also possible and contemplated.
Fig. 2 is a block diagram of a credit rating apparatus based on multi-classifier integration according to an embodiment of the present invention, and as shown in fig. 2, the credit rating apparatus based on multi-classifier integration according to an embodiment of the present invention includes: a financial characteristic data determination unit 1, a multi-classifier credit rating unit 2 and a bond issuer credit rating determination unit 3.
And the financial characteristic data determining unit 1 is used for acquiring the financial characteristic data of the bond issuer.
In an alternative embodiment of the invention, the financial characteristic data comprises: an operating revenue growth rate of four quarters, a net profit growth rate of four quarters after the unusual profit is deducted, an operating revenue to fixed asset ratio, and a flowing asset to flowing liability ratio.
And the multi-classifier credit rating unit 2 is used for inputting the financial characteristic data into a trained naive Bayes classifier model, a gradient boosting decision tree classifier model and a random forest classifier model respectively to obtain credit rating results output by the classifier models, wherein the naive Bayes classifier model, the gradient boosting decision tree classifier model and the random forest classifier model are obtained by training according to historical financial characteristic data and historical credit rating data.
And the credit rating determination unit 3 is used for determining the credit rating of the bond issuer according to the credit rating result output by each classifier model.
In an alternative embodiment of the present invention, the inventive credit rating apparatus based on multi-classifier integration further comprises: the naive Bayes classifier model training unit is used for acquiring historical financial characteristic data and historical credit rating data; and training the naive Bayes classifier model according to historical financial characteristic data, historical credit rating data and a preset naive Bayes classification algorithm, and performing credit rating on the bond issuer according to the trained naive Bayes classifier model.
In an alternative embodiment of the present invention, the inventive credit rating apparatus based on multi-classifier integration further comprises: the gradient lifting decision tree classifier model training unit is used for acquiring historical financial characteristic data and historical credit rating data; and training the gradient lifting decision tree classifier model according to historical financial characteristic data, historical credit rating data and a preset gradient lifting decision tree algorithm, so as to perform credit rating on the bond issuer according to the trained gradient lifting decision tree classifier model.
In an alternative embodiment of the present invention, the inventive credit rating apparatus based on multi-classifier integration further comprises: the random forest classifier model training unit is used for acquiring historical financial characteristic data and historical credit rating data; and training the random forest classifier model according to historical financial characteristic data, historical credit rating data and a preset random forest algorithm, and performing credit rating on the bond issuer according to the trained random forest classifier model.
To achieve the above object, according to another aspect of the present application, there is also provided a computer apparatus. As shown in fig. 3, the computer device comprises a memory, a processor, a communication interface and a communication bus, wherein a computer program that can be run on the processor is stored in the memory, and the steps of the method of the above embodiment are realized when the processor executes the computer program.
The processor may be a Central Processing Unit (CPU). The Processor may also be other general purpose processors, Digital Signal Processors (DSPs), Application Specific Integrated Circuits (ASICs), Field Programmable Gate Arrays (FPGAs) or other Programmable logic devices, discrete Gate or transistor logic devices, discrete hardware components, or a combination thereof.
The memory, which is a non-transitory computer readable storage medium, may be used to store non-transitory software programs, non-transitory computer executable programs, and units, such as the corresponding program units in the above-described method embodiments of the present invention. The processor executes various functional applications of the processor and the processing of the work data by executing the non-transitory software programs, instructions and modules stored in the memory, that is, the method in the above method embodiment is realized.
The memory may include a storage program area and a storage data area, wherein the storage program area may store an operating system, an application program required for at least one function; the storage data area may store data created by the processor, and the like. Further, the memory may include high speed random access memory, and may also include non-transitory memory, such as at least one disk storage device, flash memory device, or other non-transitory solid state storage device. In some embodiments, the memory optionally includes memory located remotely from the processor, and such remote memory may be coupled to the processor via a network. Examples of such networks include, but are not limited to, the internet, intranets, local area networks, mobile communication networks, and combinations thereof.
The one or more units are stored in the memory and when executed by the processor perform the method of the above embodiments.
The specific details of the computer device may be understood by referring to the corresponding related descriptions and effects in the above embodiments, and are not described herein again.
In order to achieve the above object, according to another aspect of the present application, there is also provided a computer-readable storage medium storing a computer program which, when executed in a computer processor, implements the steps in the above-mentioned credit rating method based on multi-classifier integration. It will be understood by those skilled in the art that all or part of the processes of the methods of the embodiments described above can be implemented by a computer program, which can be stored in a computer-readable storage medium, and when executed, can include the processes of the embodiments of the methods described above. The storage medium may be a magnetic Disk, an optical Disk, a Read-Only Memory (ROM), a Random Access Memory (RAM), a Flash Memory (Flash Memory), a Hard Disk (Hard Disk Drive, abbreviated as HDD) or a Solid State Drive (SSD), etc.; the storage medium may also comprise a combination of memories of the kind described above.
It will be apparent to those skilled in the art that the modules or steps of the present invention described above may be implemented by a general purpose computing device, they may be centralized on a single computing device or distributed across a network of multiple computing devices, and they may alternatively be implemented by program code executable by a computing device, such that they may be stored in a storage device and executed by a computing device, or fabricated separately as individual integrated circuit modules, or fabricated as a single integrated circuit module from multiple modules or steps. Thus, the present invention is not limited to any specific combination of hardware and software.
The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention, and various modifications and changes may be made by those skilled in the art. Any modification, equivalent replacement, or improvement made within the spirit and principle of the present invention should be included in the protection scope of the present invention.

Claims (12)

1. A credit rating method based on multi-classifier integration, comprising:
acquiring financial characteristic data of a bond issuer;
respectively inputting the financial characteristic data into a trained naive Bayes classifier model, a gradient boosting decision tree classifier model and a random forest classifier model to obtain credit rating results output by the classifier models, wherein the naive Bayes classifier model, the gradient boosting decision tree classifier model and the random forest classifier model are obtained by training according to historical financial characteristic data and historical credit rating data;
and determining the credit rating of the bond issuer according to the credit rating result output by each classifier model.
2. The multi-classifier integration-based credit rating method of claim 1, wherein the financial characteristic data comprises: an operating revenue growth rate of four quarters, a net profit growth rate of four quarters after the unusual profit is deducted, an operating revenue to fixed asset ratio, and a flowing asset to flowing liability ratio.
3. The multi-classifier integration-based credit rating method of claim 1 or 2, further comprising:
acquiring historical financial characteristic data and historical credit rating data;
training the naive Bayes classifier model according to historical financial feature data, historical credit rating data and a preset naive Bayes classification algorithm, and performing credit rating on the bond issuer according to the trained naive Bayes classifier model.
4. The multi-classifier integration-based credit rating method of claim 1 or 2, further comprising:
acquiring historical financial characteristic data and historical credit rating data;
training the gradient lifting decision tree classifier model according to historical financial feature data, historical credit rating data and a preset gradient lifting decision tree algorithm, and performing credit rating on the bond issuer according to the trained gradient lifting decision tree classifier model.
5. The multi-classifier integration-based credit rating method of claim 1 or 2, further comprising:
acquiring historical financial characteristic data and historical credit rating data;
training the random forest classifier model according to historical financial characteristic data, historical credit rating data and a preset random forest algorithm, and performing credit rating on a bond issuer according to the trained random forest classifier model.
6. A credit rating apparatus based on multi-classifier integration, comprising:
the financial characteristic data determining unit is used for acquiring financial characteristic data of the bond issuer;
the multi-classifier credit rating unit is used for respectively inputting the financial characteristic data into a trained naive Bayes classifier model, a gradient boosting decision tree classifier model and a random forest classifier model to obtain credit rating results output by the classifier models, wherein the naive Bayes classifier model, the gradient boosting decision tree classifier model and the random forest classifier model are obtained by training according to historical financial characteristic data and historical credit rating data;
and the credit rating determination unit of the bond issuer is used for determining the credit rating of the bond issuer according to the credit rating result output by each classifier model.
7. The multi-classifier integration-based credit rating device of claim 6, wherein the financial characteristic data comprises: an operating revenue growth rate of four quarters, a net profit growth rate of four quarters after the unusual profit is deducted, an operating revenue to fixed asset ratio, and a flowing asset to flowing liability ratio.
8. The multi-classifier integration-based credit rating device of claim 6 or 7, further comprising:
the naive Bayes classifier model training unit is used for acquiring historical financial characteristic data and historical credit rating data; and training the naive Bayes classifier model according to historical financial characteristic data, historical credit rating data and a preset naive Bayes classification algorithm, and performing credit rating on the bond issuer according to the trained naive Bayes classifier model.
9. The multi-classifier integration-based credit rating device of claim 6 or 7, further comprising:
the gradient lifting decision tree classifier model training unit is used for acquiring historical financial characteristic data and historical credit rating data; and training the gradient lifting decision tree classifier model according to historical financial characteristic data, historical credit rating data and a preset gradient lifting decision tree algorithm, so as to perform credit rating on the bond issuer according to the trained gradient lifting decision tree classifier model.
10. The multi-classifier integration-based credit rating device of claim 6 or 7, further comprising:
the random forest classifier model training unit is used for acquiring historical financial characteristic data and historical credit rating data; and training the random forest classifier model according to historical financial characteristic data, historical credit rating data and a preset random forest algorithm, and performing credit rating on the bond issuer according to the trained random forest classifier model.
11. A computer device comprising a memory, a processor and a computer program stored on the memory and executable on the processor, characterized in that the processor implements the method of any of claims 1 to 5 when executing the computer program.
12. A computer-readable storage medium, in which a computer program is stored which, when executed in a computer processor, implements the method of any one of claims 1 to 5.
CN202010009078.5A 2020-01-06 2020-01-06 Credit rating method and device based on multi-classifier integration Pending CN111242191A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202010009078.5A CN111242191A (en) 2020-01-06 2020-01-06 Credit rating method and device based on multi-classifier integration

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202010009078.5A CN111242191A (en) 2020-01-06 2020-01-06 Credit rating method and device based on multi-classifier integration

Publications (1)

Publication Number Publication Date
CN111242191A true CN111242191A (en) 2020-06-05

Family

ID=70872449

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202010009078.5A Pending CN111242191A (en) 2020-01-06 2020-01-06 Credit rating method and device based on multi-classifier integration

Country Status (1)

Country Link
CN (1) CN111242191A (en)

Cited By (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN111782807A (en) * 2020-06-19 2020-10-16 西北工业大学 Self-acceptance technology debt detection and classification method based on multi-method ensemble learning
CN112529695A (en) * 2020-12-23 2021-03-19 招联消费金融有限公司 Credit risk determination method, credit risk determination device, computer equipment and storage medium
CN112819069A (en) * 2021-01-29 2021-05-18 中国农业银行股份有限公司 Event grading method and device

Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN106803204A (en) * 2017-01-17 2017-06-06 上海聚赟大数据有限公司 A kind of floater's promise breaking analysis method based on machine learning
CN107248030A (en) * 2017-05-26 2017-10-13 谢首鹏 A kind of bond Risk Forecast Method and system based on machine learning algorithm
CN107766418A (en) * 2017-09-08 2018-03-06 广州汪汪信息技术有限公司 A kind of credit estimation method based on Fusion Model, electronic equipment and storage medium
CN110490726A (en) * 2019-07-11 2019-11-22 林宏伟 Credit debt risk management method, device, equipment and storage medium
CN110503459A (en) * 2019-07-19 2019-11-26 平安科技（深圳）有限公司 User credit degree appraisal procedure, device and storage medium based on big data
CN110580268A (en) * 2019-08-05 2019-12-17 西北大学 Credit scoring integrated classification system and method based on deep learning

Patent Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN106803204A (en) * 2017-01-17 2017-06-06 上海聚赟大数据有限公司 A kind of floater's promise breaking analysis method based on machine learning
CN107248030A (en) * 2017-05-26 2017-10-13 谢首鹏 A kind of bond Risk Forecast Method and system based on machine learning algorithm
CN107766418A (en) * 2017-09-08 2018-03-06 广州汪汪信息技术有限公司 A kind of credit estimation method based on Fusion Model, electronic equipment and storage medium
CN110490726A (en) * 2019-07-11 2019-11-22 林宏伟 Credit debt risk management method, device, equipment and storage medium
CN110503459A (en) * 2019-07-19 2019-11-26 平安科技（深圳）有限公司 User credit degree appraisal procedure, device and storage medium based on big data
CN110580268A (en) * 2019-08-05 2019-12-17 西北大学 Credit scoring integrated classification system and method based on deep learning

Non-Patent Citations (2)

* Cited by examiner, † Cited by third party
Title
李艳秋: "基于集成学习的人脸识别研究", 《中国博士学位论文全文数据库 信息科技辑》, no. 02, 15 February 2019 (2019-02-15), pages 45 - 102 *
王冲: "《现代信息检索技术基本原理教程》", 西安电子科技大学出版社, pages: 98 - 102 *

Cited By (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN111782807A (en) * 2020-06-19 2020-10-16 西北工业大学 Self-acceptance technology debt detection and classification method based on multi-method ensemble learning
CN111782807B (en) * 2020-06-19 2024-05-24 西北工业大学 Self-bearing technology debt detection classification method based on multiparty integrated learning
CN112529695A (en) * 2020-12-23 2021-03-19 招联消费金融有限公司 Credit risk determination method, credit risk determination device, computer equipment and storage medium
CN112819069A (en) * 2021-01-29 2021-05-18 中国农业银行股份有限公司 Event grading method and device

Similar Documents

Publication Publication Date Title
Xia et al. Cost-sensitive boosted tree for loan evaluation in peer-to-peer lending
CN108475393A (en) The system and method that decision tree is predicted are promoted by composite character and gradient
CN111242191A (en) Credit rating method and device based on multi-classifier integration
WO2019062006A1 (en) Time selection admission method based on machine learning, device and terminal equipment therefor
CN111967971B (en) Bank customer data processing method and device
CN108898476A (en) A kind of loan customer credit-graded approach and device
CN113095927B (en) Method and equipment for identifying suspected transactions of backwashing money
KR102342321B1 (en) Method and apparatus for deep learning-based stock screening and its use for advancing and automating portfolio management
Song Stock trend prediction: Based on machine learning methods
Khodaee et al. Forecasting turning points in stock price by applying a novel hybrid CNN-LSTM-ResNet model fed by 2D segmented images
Cheng et al. A seasonal time-series model based on gene expression programming for predicting financial distress
CN112613997A (en) Method and apparatus for forecasting combined investment of money fund
CN112434862B (en) Method and device for predicting financial dilemma of marketing enterprises
Sutkatti et al. Stock market forecasting techniques: A survey
CN108629381A (en) Crowd's screening technique based on big data and terminal device
CN115186898A (en) Bank account opening amount prediction method and device
CN111667307B (en) Method and device for predicting financial product sales volume
Janosek et al. Knowledge discovery in dynamic data using neural networks
CN109308565B (en) Crowd performance grade identification method and device, storage medium and computer equipment
CN111340632A (en) Resource control method and device
Borovkova et al. Deep learning prediction of the eurostoxx 50 with news sentiment
Dhivya et al. Stock Price Prediction based on Generative Adversarial Network
Ji et al. Predictive Analysis of Cryptocurrency Residual Log-returns Based on Lightgbm
Wanat et al. In Search of Hedges and Safe Havens in Global Financial Markets
Nayak et al. Predicting Stock Market Price Using Machine Learning Techniques

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
TA01 Transfer of patent application right

Effective date of registration: 20220908

Address after: 25 Financial Street, Xicheng District, Beijing 100033

Applicant after: CHINA CONSTRUCTION BANK Corp.

Address before: 25 Financial Street, Xicheng District, Beijing 100033

Applicant before: CHINA CONSTRUCTION BANK Corp.

Applicant before: Jianxin Financial Science and Technology Co.,Ltd.

TA01 Transfer of patent application right