CN108090776A - The update method and system of credit evaluation model, credit estimation method and system - Google Patents
The update method and system of credit evaluation model, credit estimation method and system Download PDFInfo
- Publication number
- CN108090776A CN108090776A CN201711327015.9A CN201711327015A CN108090776A CN 108090776 A CN108090776 A CN 108090776A CN 201711327015 A CN201711327015 A CN 201711327015A CN 108090776 A CN108090776 A CN 108090776A
- Authority
- CN
- China
- Prior art keywords
- model
- credit evaluation
- evaluation model
- new
- index
- 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
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
- G06Q30/00—Commerce
- G06Q30/018—Certifying business or products
-
- 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
- G06Q30/00—Commerce
- G06Q30/06—Buying, selling or leasing transactions
- G06Q30/0601—Electronic shopping [e-shopping]
- G06Q30/0609—Buyer or seller confidence or verification
Abstract
The present invention discloses a kind of update method for being capable of credit evaluation model and more new system, credit estimation method and credit evaluation system.Update method includes：Obtain new credit evaluation model input by user；The Model Matching degree of new credit evaluation model is determined according to the effective information, model test coverage and model accuracy of new credit evaluation model；Judge whether the Model Matching degree of new credit evaluation model is more than the Model Matching degree of former credit evaluation model, obtain the first judging result；When the Model Matching that the Model Matching degree of new credit evaluation model is more than former credit evaluation model is spent, former credit evaluation model is replaced with new credit evaluation model；When the Model Matching that the Model Matching degree of new credit evaluation model is less than or equal to former credit evaluation model is spent, retain former credit evaluation model.The accuracy of credit evaluation can be improved using update method provided by the invention and more new system, credit estimation method and credit evaluation system.
Description
Technical field
The present invention relates to e-commerce field, update method and system more particularly to a kind of credit evaluation model, letters
With appraisal procedure and system.
Background technology
At present, common problem existing for all kinds of electronic commerce credits assessments is：The correlation of credit can not be ensured to assess
Data it is true, accurate；The participation of the most someone of evaluation process usually has the composition that gets sth into one's head, and can not ensure the standard of assessment
True property.Meanwhile each credit evaluation model or evaluation criteria disunity, and without unified more new standard.Over time,
When carrying out credit evaluation to gate system, the assessment rule for having many assessment models is no longer applicable in, if do not updated assessment mould
Type will substantially reduce the accuracy of assessment.
Therefore, the technical issues of how improving the accuracy of credit evaluation, becoming those skilled in the art's urgent need to resolve.
The content of the invention
The object of the present invention is to provide a kind of update method of credit evaluation model and more new system, credit estimation method and
Credit evaluation system can improve the accuracy of credit evaluation.
To achieve the above object, the present invention provides following schemes：
A kind of update method of credit evaluation model, the update method include：
Obtain new credit evaluation model input by user；
The new letter is determined according to effective information, model test coverage and the model accuracy of the new credit evaluation model
With the Model Matching degree of assessment models；
Judge whether the Model Matching degree of the new credit evaluation model is more than the Model Matching degree of former credit evaluation model,
Obtain the first judging result；
When the Model Matching that the Model Matching degree of the new credit evaluation model is more than the former credit evaluation model is spent,
The former credit evaluation model is replaced with the new credit evaluation model；
When the Model Matching degree of the new credit evaluation model is less than or equal to the model of the former credit evaluation model
During matching degree, retain the former credit evaluation model.
Optionally, effective information, model test coverage and the model accuracy according to the new credit evaluation model
It determines the Model Matching degree of the new credit evaluation model, specifically includes：
The new credit evaluation model is determined according to the Information quanlity index of each index of the new credit evaluation model
Effective information；
The model of the new credit evaluation model is determined according to the coverage rate of each index of the new credit evaluation model
Coverage rate；
The model accuracy of the new credit evaluation model is determined according to each root-mean-square error, wherein, the root mean square
Error is the root-mean-square error of assessment result when being assessed using the new credit evaluation model sample set；
The new credit evaluation mould is calculated according to the effective information, the model test coverage and the model accuracy
The Model Matching degree of type.
Optionally, the Information quanlity index of each index according to the new credit evaluation model determines the new credit
The effective information of assessment models, specifically includes：
According to formula：Determine i-th of index of the new credit evaluation model
Information quanlity index, wherein, HiRepresent the Information quanlity index of i-th of index, xijRepresent i-th of index it is corresponding j-th it is single
Event, χiRepresent the corresponding event sets of i-th of index, p (xij) represent xijThe probability of generation；
According to formula：Determine the effective information of the new credit evaluation model, wherein, V represents effectively letter
Breath amount, N represent index quantity.
Optionally, the model accuracy that the new credit evaluation model is determined according to each root-mean-square error, specifically
Including：
According to formula：Determine the corresponding root-mean-square error of each sample set,
In, erriRepresent the corresponding root-mean-square error of i-th of sample set, M represents the sample size in i-th of sample set, scoreijTable
Show fractional value when the new credit evaluation model assesses j-th of sample in i-th of sample set,Represent i-th of sample
The standard score for j-th of the sample concentrated；
According to formula：Determine the model accuracy of the new credit evaluation model, wherein, Q is represented
Model accuracy, the quantity of N ' expression sample sets, piRepresent the norm of the corresponding root-mean-square error of i-th of sample set.
A kind of more new system of credit evaluation model, the more new system include：
New model acquisition module, for obtaining new credit evaluation model input by user；
Matching degree determining module, for effective information, model test coverage and the mould according to the new credit evaluation model
Type accuracy determines the Model Matching degree of the new credit evaluation model；
First judgment module, for judging whether the Model Matching degree of the new credit evaluation model is more than former credit evaluation
The Model Matching degree of model obtains the first judging result；
Processing module is more than the former credit evaluation model for working as the Model Matching degree of the new credit evaluation model
When Model Matching is spent, the former credit evaluation model is replaced with the new credit evaluation model；
When the Model Matching degree of the new credit evaluation model is less than or equal to the model of the former credit evaluation model
During matching degree, retain the former credit evaluation model.
Optionally, the matching degree computing module specifically includes：
Effective information determination unit, the Information quanlity index for each index according to the new credit evaluation model are true
The effective information of the fixed new credit evaluation model；
Model test coverage determination unit, the coverage rate for each index according to the new credit evaluation model determine institute
State the model test coverage of new credit evaluation model；
Model accuracy determination unit, for determining the model of the new credit evaluation model according to each root-mean-square error
Accuracy, wherein, the root-mean-square error is assessment result when being assessed using the new credit evaluation model sample set
Root-mean-square error；
Model Matching degree computing unit, for accurate according to the effective information, the model test coverage and the model
Exactness calculates the Model Matching degree of the new credit evaluation model.
Optionally, the effective information determination unit specifically includes：
Information quanlity index determination subelement, for according to formula：It determines described
The Information quanlity index of i-th of index of new credit evaluation model, wherein, HiRepresent the Information quanlity index of i-th of index, xijIt represents
Corresponding j-th of the single incident of i-th of index, χiRepresent the corresponding event sets of i-th of index, p (xij) represent xijOccur
Probability；
Effective information determination subelement, for according to formula：Determine having for the new credit evaluation model
Information content is imitated, wherein, V represents effective information, and N represents index quantity.
Optionally, the model accuracy determination unit specifically includes：
Error determination subelement, for according to formula：Determine each sample set
Corresponding root-mean-square error, wherein, erriRepresent the corresponding root-mean-square error of i-th of sample set, M is represented in i-th of sample set
Sample size, scoreijRepresent fraction when the new credit evaluation model assesses j-th of sample in i-th of sample set
Value,Represent the standard score of j-th of sample in i-th of sample set；
Accuracy determination subelement, for according to formula：Determine the new credit evaluation model
Model accuracy, wherein, Q represents model accuracy, the quantity of N ' expression sample sets, piRepresent that i-th of sample set is corresponding
The norm of square error.
A kind of credit estimation method based on block chain, the block chain include multiple gate systems and multiple data
Chain, each gate system correspond to a data subchain, each data subchain for record in chronological order with
Commercial matters information data, government affairs information data and the social information data of the corresponding gate system of the data subchain；The credit
Appraisal procedure includes：
The credit data of gate system to be assessed is obtained, the credit data is corresponding with the gate system to be assessed
The commercial matters information data, the government affairs information data and/or the social information data recorded in data subchain；
The credit evaluation value of the gate system to be assessed is determined according to the credit data and credit evaluation model,
In, the credit evaluation model is the updated credit evaluation model of update method.
A kind of credit evaluation system based on block chain, the block chain include multiple gate systems and multiple data
Chain, each gate system correspond to a data subchain, each data subchain for record in chronological order with
Commercial matters information data, government affairs information data and the social information data of the corresponding gate system of the data subchain；The credit
Assessment system includes：
Credit data acquisition module, for obtaining the credit data of gate system to be assessed, the credit data for institute
State the commercial matters information data recorded in the corresponding data subchain of gate system to be assessed, the government affairs information data and/or
Social information data；
Evaluation module, for determining the letter of the gate system to be assessed according to the credit data and credit evaluation model
With assessed value, wherein, the credit evaluation model is the updated credit evaluation model of update method.
The specific embodiment provided according to the present invention, the invention discloses following technique effects：
The present invention determines institute according to the effective information, model test coverage and model accuracy of new credit evaluation model first
The Model Matching degree of new credit evaluation model is stated, then according to the Model Matching degree of new credit evaluation model and former credit evaluation mould
The magnitude relationship of the Model Matching degree of type determines whether to update credit evaluation model.When the Model Matching degree of new credit evaluation model
When Model Matching more than former credit evaluation model is spent, former credit evaluation model is replaced with new credit evaluation model, otherwise, is retained
Former credit evaluation model, thereby it is ensured that the Model Matching degree higher of the assessment models of credit evaluation is carried out to gate system,
So as to improve the accuracy of credit evaluation.
The present invention is data subchain corresponding with gate system to be assessed for assessing the credit data of gate system credit
Commercial matters information data, government affairs information data and/or the social information data of middle record, it can be ensured that for assessing the correlation of credit
The authenticities of data, validity, objectivity and comprehensive, and assessed using the higher assessment models of Model Matching degree, it is whole
A evaluation process carries out automatically, is participated in without artificial, therefore can eliminate the composition that gets sth into one's head, and further improves the accurate of assessment
Property.
Description of the drawings
It in order to illustrate more clearly about the embodiment of the present invention or technical scheme of the prior art, below will be to institute in embodiment
Attached drawing to be used is needed to be briefly described, it should be apparent that, the accompanying drawings in the following description is only some implementations of the present invention
Example, for those of ordinary skill in the art, without creative efforts, can also obtain according to these attached drawings
Obtain other attached drawings.
Fig. 1 is the flow chart for the update method that the embodiment of the present invention 1 provides；
Fig. 2 is the structure diagram for the more new system that the embodiment of the present invention 2 provides.
Specific embodiment
Below in conjunction with the attached drawing in the embodiment of the present invention, the technical solution in the embodiment of the present invention is carried out clear, complete
Site preparation describes, it is clear that described embodiment is only part of the embodiment of the present invention, instead of all the embodiments.It is based on
Embodiment in the present invention, those of ordinary skill in the art are obtained every other without making creative work
Embodiment belongs to the scope of protection of the invention.
The object of the present invention is to provide a kind of update method of credit evaluation model and more new system, credit estimation method and
Credit evaluation system can improve the accuracy of credit evaluation.
In order to make the foregoing objectives, features and advantages of the present invention clearer and more comprehensible, it is below in conjunction with the accompanying drawings and specific real
Applying mode, the present invention is described in further detail.
Embodiment 1：
Fig. 1 is the flow chart for the update method that the embodiment of the present invention 1 provides.As shown in Figure 1, a kind of credit evaluation model
Update method, the update method include：
Step 11：Obtain new credit evaluation model input by user；
Step 12：It is determined according to effective information, model test coverage and the model accuracy of the new credit evaluation model
The Model Matching degree of the new credit evaluation model；
Step 13：Judge whether the Model Matching degree of the new credit evaluation model is more than the model of former credit evaluation model
Matching degree obtains the first judging result；
When the Model Matching that the Model Matching degree of the new credit evaluation model is more than the former credit evaluation model is spent,
Perform step 14；
When the Model Matching degree of the new credit evaluation model is less than or equal to the model of the former credit evaluation model
During matching degree, step 15 is performed；
Step 14：The former credit evaluation model is replaced with the new credit evaluation model；
Step 15：Retain the former credit evaluation model.
Specifically, step 12：Effective information, model test coverage and the model according to the new credit evaluation model
Accuracy determines the Model Matching degree of the new credit evaluation model, specifically includes：
Step 121：Determine that the new credit is commented according to the Information quanlity index of each index of the new credit evaluation model
Estimate the effective information of model；
Step 122：The new credit evaluation mould is determined according to the coverage rate of each index of the new credit evaluation model
The model test coverage of type.
Model test coverage refers to that the evaluation index that model uses covers the coverage rate of all door available informations, can basis
Formula：Computation model coverage rate, wherein CiRepresent the coverage rate of i-th of index, N represents the quantity of index, all fingers
The sum of target coverage rate is total coverage rate of model.
In the present embodiment, evaluation index is divided into multiple levels, the index of each level divides equally current level remaining proportion,
If the index has next level, which transfers next level by 50% coverage rate index.For example, certain gate system is gathered around
There are top index a, b, a c, secondary index six aa, ab, ac, ba, bb, ca, if model includes aa, ba, bb, tetra- indexs of ca,
Then index coverage rate calculation is：C (a)=C (b)=C (c)=1/6, C (aa)=C (ab)=C (ac)=1/18, C (ba)
=C (bb)=1/12, C (ca)=1/6.Model test coverage calculation is C (a)+C (aa)+C (b)+C (ba)+C (bb)+C (c)
+ C (ca)=88.9%.
Step 123：The model accuracy of the new credit evaluation model is determined according to each root-mean-square error, wherein, institute
State the root-mean-square error that root-mean-square error is assessment result when being assessed using the new credit evaluation model sample set；
Step 124：The new letter is calculated according to the effective information, the model test coverage and the model accuracy
With the Model Matching degree of assessment models.
Wherein, step 121：Described in the Information quanlity index of each index according to the new credit evaluation model determines
The effective information of new credit evaluation model, effective information refer to the sum of Information quanlity index of index that the model uses,
It specifically includes：
Step 1211：According to comentropy formula：Determine the new credit evaluation
The Information quanlity index of i-th of index of model, wherein, HiRepresent the Information quanlity index of i-th of index, xijRepresent i-th of index
Corresponding j-th of single incident, χiRepresent the corresponding event sets of i-th of index, p (xij) represent xijThe probability of generation；
Step 1212：According to formula：Determine the effective information of the new credit evaluation model, wherein, V tables
Show effective information, N represents index quantity.
Step 123：The model accuracy that the new credit evaluation model is determined according to each root-mean-square error, specifically
Including：
Step 1231：According to formula：Determine that each sample set is corresponding square
Root error, wherein, erriRepresenting the corresponding root-mean-square error of i-th of sample set, M represents the sample size in i-th of sample set,
scoreijRepresent fractional value when the new credit evaluation model assesses j-th of sample in i-th of sample set,It represents
The standard score of j-th of sample in i-th of sample set, standard score can be provided by specialty evaluation mechanism.When more new model,
Specialty evaluation mechanism can be according to performance of the model under extensive sample come test model accuracy, to prevent over-fitting etc..
Step 1232：According to formula：Determine the model accuracy of the new credit evaluation model,
In, Q represents model accuracy, the quantity of N ' expression sample sets, piRepresent the model of the corresponding root-mean-square error of i-th of sample set
Number.
Specifically, step 12：It is accurate according to the effective information, model test coverage and model of the new credit evaluation model
Degree determines the Model Matching degree of the new credit evaluation model, specifically includes:
According to formula：D=ω1V+ω2C+ω3* Q determines the Model Matching degree of the new credit evaluation model, wherein, D
Represent Model Matching degree, ω1Represent effective information coefficient of discharge, ω2Represent model test coverage coefficient, ω2Represent model accuracy system
Number, C represent model test coverage.
The present embodiment determines first according to the effective information, model test coverage and model accuracy of new credit evaluation model
The Model Matching degree of the new credit evaluation model, then according to the Model Matching degree of new credit evaluation model and former credit evaluation
The magnitude relationship of the Model Matching degree of model determines whether to update credit evaluation model.When the Model Matching of new credit evaluation model
When the Model Matching that degree is more than former credit evaluation model is spent, former credit evaluation model is replaced with new credit evaluation model, otherwise, is protected
Former credit evaluation model is stayed, thereby it is ensured that carrying out the Model Matching degree of the assessment models of credit evaluation to gate system more
Height, so as to improve the accuracy of credit evaluation.
Embodiment 2：
Fig. 2 is the structure diagram for the more new system that the embodiment of the present invention 2 provides.As shown in Fig. 2, a kind of credit evaluation model
More new system, the more new system includes：
New model acquisition module 21, for obtaining new credit evaluation model input by user；
Matching degree determining module 22, for according to the new credit evaluation model effective information, model test coverage and
Model accuracy determines the Model Matching degree of the new credit evaluation model；
First judgment module 23 is commented for judging whether the Model Matching degree of the new credit evaluation model is more than former credit
Estimate the Model Matching degree of model, obtain the first judging result；
Processing module 24 is more than the former credit evaluation model for working as the Model Matching degree of the new credit evaluation model
Model Matching when spending, replace the former credit evaluation model with the new credit evaluation model；
When the Model Matching degree of the new credit evaluation model is less than or equal to the model of the former credit evaluation model
During matching degree, retain the former credit evaluation model.
Specifically, the matching degree computing module 22 specifically includes：
Effective information determination unit 221, the information content for each index according to the new credit evaluation model refer to
Number determines the effective information of the new credit evaluation model；
Model test coverage determination unit 222, the coverage rate for each index according to the new credit evaluation model are true
The model test coverage of the fixed new credit evaluation model；
Model accuracy determination unit 223, for determining the new credit evaluation model according to each root-mean-square error
Model accuracy, wherein, the root-mean-square error is assessed when being and sample set being assessed using the new credit evaluation model
As a result root-mean-square error；
Model Matching degree computing unit 224, for according to the effective information, the model test coverage and the model
Accuracy calculates the Model Matching degree of the new credit evaluation model.
Further, the effective information determination unit 221 specifically includes：
Information quanlity index determination subelement 2211, for according to formula：It determines
The Information quanlity index of i-th of index of the new credit evaluation model, wherein, HiRepresent the Information quanlity index of i-th of index, xij
Represent corresponding j-th of the single incident of i-th of index, χiRepresent the corresponding event sets of i-th of index, p (xij) represent xijHair
Raw probability；
Effective information determination subelement 2212, for according to formula：Determine the new credit evaluation model
Effective information, wherein, V represent effective information, N represent index quantity.
The model accuracy determination unit 223 specifically includes：
Error determination subelement 2231, for according to formula：Determine each sample
The corresponding root-mean-square error of this collection, wherein, erriRepresent the corresponding root-mean-square error of i-th of sample set, M represents i-th of sample
The sample size of concentration, scoreijRepresent point when the new credit evaluation model assesses j-th of sample in i-th of sample set
Numerical value,Represent the standard score of j-th of sample in i-th of sample set；
Accuracy determination subelement 2232, for according to formula：Determine the new credit evaluation mould
The model accuracy of type, wherein, Q represents model accuracy, the quantity of N ' expression sample sets, piRepresent that i-th of sample set corresponds to
Root-mean-square error norm.
The present embodiment is according to the Model Matching degree of new credit evaluation model and the Model Matching degree of former credit evaluation model
Magnitude relationship determines whether to update credit evaluation model, when carrying out credit evaluation to gate system, the Model Matching of assessment models
Spend higher, therefore the accuracy higher of credit evaluation.
Embodiment 3：
A kind of credit estimation method based on block chain, the block chain include multiple gate systems and multiple data
Chain, each gate system correspond to a data subchain, each data subchain for record in chronological order with
Commercial matters information data, government affairs information data and the social information data of the corresponding gate system of the data subchain；The credit
Appraisal procedure includes：
Step 31：The credit data of gate system to be assessed is obtained, the credit data is and the door system to be assessed
Commercial matters information data, government affairs information data and/or the social information data recorded in corresponding data subchain of uniting；
Step 32：The credit evaluation of the gate system to be assessed is determined according to the credit data and credit evaluation model
Value, wherein, the credit evaluation model is the updated credit evaluation model of update method according to embodiment 1.
Holographyization gate system block chain in the present embodiment is that have commercial affairs, political affairs by gate system to be assessed and with it
Business, correlation individual, enterprise and the mechanism gate system of social networks are formed.Holographyization gate system is that holographyization is personal, looks forward to
Industry and mechanism cyberspace carrier and the form of expression, by " I " component, " my demand " component, " my supply " component,
" my space " component is formed, and the basic information of description individual, enterprise and mechanism, demand information, supply information, society close respectively
It is information.Holographyization gate system block chain is made of two subchains, and one is door subchain, and one is data subchain, two
There are one common wound generation blocks for subchain tool.Wound generation block is used to record the creation time of holographyization gate system, network identity, its institute
The individual of representative, the basic information of enterprise or mechanism, demand information, supply information.Door subchain is used to record in temporal sequence
Add in the related gate system of the block chain, the addition time of each block record correlation gate system therein, network identity
And the intelligent contract with the gate system.Data subchain is used to record in temporal sequence between the gate system and related gate system
All relevant commercial affairs, government affairs and social informations.
In the present embodiment, for assessing the credit data of gate system credit as data corresponding with gate system to be assessed
Commercial matters information data, government affairs information data and/or the social information data recorded in subchain, it can be ensured that for assessing credit
The authenticity of related data, validity and comprehensive, and assessed using the higher assessment models of Model Matching degree, entirely comment
Estimate process to carry out automatically, be participated in without artificial, therefore the composition that gets sth into one's head can be eliminated, further improve the accuracy of assessment.
Embodiment 4：
A kind of credit evaluation system based on block chain, the block chain include multiple gate systems and multiple data
Chain, each gate system correspond to a data subchain, each data subchain for record in chronological order with
Commercial matters information data, government affairs information data and the social information data of the corresponding gate system of the data subchain；The credit
Assessment system includes：
Credit data acquisition module 41, for obtaining the credit data of gate system to be assessed, the credit data be with
The commercial matters information data that are recorded in the corresponding data subchain of the gate system to be assessed, the government affairs information data and/
Or social information data；
Evaluation module 42, for determining the gate system to be assessed according to the credit data and credit evaluation model
Credit evaluation value, wherein, the credit evaluation model is the updated credit evaluation model of update method according to embodiment 1.
The credit data of the present embodiment be the corresponding data subchain of gate system to be assessed in record commercial matters information data,
Government affairs information data and/or social information data, it can be ensured that for assess the authenticity of the related data of credit, validity and
It is comprehensive, and assessed using the higher assessment models of Model Matching degree, entire evaluation process carries out automatically, without artificial ginseng
With, therefore the composition that gets sth into one's head can be eliminated, further improve the accuracy of assessment.
Credit automatic evaluation system provided in this embodiment based on holographyization gate system block chain is using the continuous, cycle
Property evaluation profile, i.e. base period credit evaluation forms base period credit evaluation as a result, follow-up credit evaluation result is after being automatically performed
The credit related data in assessment cycle is gathered, is formed according to standardization credit evaluation model through automatic assessment in assessment cycle
Credit evaluation is updated credit early period as a result, the credit evaluation result is based on credit update rule, forms current credit.
Credit automatic evaluation system provided in this embodiment based on holographyization gate system block chain, main purpose is to build
Found a kind of credit automatic evaluation system based on holographyization gate system block chain.Its distinguishing feature is：1) ensure to assess
The related data of credit it is true, accurate, comprehensive；2) evaluation process carries out automatically, participation and appraisal agency's intervention without people,
Elimination gets sth into one's head composition；3) using standardization credit evaluation model, it is ensured that the comparativity and applicability of credit evaluation result；4)
Periodically automatic assessment is simultaneously updated by the phase, it is ensured that the continuity of credit evaluation.
Each embodiment is described by the way of progressive in this specification, the highlights of each of the examples are with other
The difference of embodiment, just to refer each other for identical similar portion between each embodiment.For system disclosed in embodiment
For, since it is corresponded to the methods disclosed in the examples, so description is fairly simple, related part is said referring to method part
It is bright.
Specific case used herein is set forth the principle of the present invention and embodiment, and above example is said
It is bright to be only intended to help the method and its core concept for understanding the present invention；Meanwhile for those of ordinary skill in the art, foundation
The thought of the present invention, in specific embodiments and applications there will be changes.In conclusion this specification content is not
It is interpreted as limitation of the present invention.
Claims (10)
1. a kind of update method of credit evaluation model, which is characterized in that the update method includes：
Obtain new credit evaluation model input by user；
Determine that the new credit is commented according to effective information, model test coverage and the model accuracy of the new credit evaluation model
Estimate the Model Matching degree of model；
Judge whether the Model Matching degree of the new credit evaluation model is more than the Model Matching degree of former credit evaluation model, obtain
First judging result；
When the Model Matching that the Model Matching degree of the new credit evaluation model is more than the former credit evaluation model is spent, institute is used
It states new credit evaluation model and replaces the former credit evaluation model；
When the Model Matching degree of the new credit evaluation model is less than or equal to the Model Matching of the former credit evaluation model
When spending, retain the former credit evaluation model.
2. update method according to claim 1, which is characterized in that described according to the effective of the new credit evaluation model
Information content, model test coverage and model accuracy determine the Model Matching degree of the new credit evaluation model, specifically include：
The effective of the new credit evaluation model is determined according to the Information quanlity index of each index of the new credit evaluation model
Information content；
The model for determining the new credit evaluation model according to the coverage rate of each index of the new credit evaluation model covers
Rate；
The model accuracy of the new credit evaluation model is determined according to each root-mean-square error, wherein, the root-mean-square error
The root-mean-square error of assessment result during to be assessed using the new credit evaluation model sample set；
The new credit evaluation model is calculated according to the effective information, the model test coverage and the model accuracy
Model Matching degree.
3. update method according to claim 2, which is characterized in that described according to each of the new credit evaluation model
The Information quanlity index of index determines the effective information of the new credit evaluation model, specifically includes：
According to formula：Determine the letter of i-th of index of the new credit evaluation model
Volume index is ceased, wherein, HiRepresent the Information quanlity index of i-th of index, xijRepresent corresponding j-th of the single incident of i-th of index,
χiRepresent the corresponding event sets of i-th of index, p (xij) represent xijThe probability of generation；
According to formula：Determine the effective information of the new credit evaluation model, wherein, V represents effective information,
N represents index quantity.
4. update method according to claim 2, which is characterized in that it is described determined according to each root-mean-square error it is described new
The model accuracy of credit evaluation model, specifically includes：
According to formula：Determine the corresponding root-mean-square error of each sample set, wherein,
erriRepresent the corresponding root-mean-square error of i-th of sample set, M represents the sample size in i-th of sample set, scoreijRepresent institute
Fractional value when new credit evaluation model assesses j-th of sample in i-th of sample set is stated,It represents in i-th of sample set
J-th of sample standard score；
According to formula：Determine the model accuracy of the new credit evaluation model, wherein, Q represents that model is accurate
Exactness, the quantity of N ' expression sample sets, piRepresent the norm of the corresponding root-mean-square error of i-th of sample set.
5. the more new system of a kind of credit evaluation model, which is characterized in that the more new system includes：
New model acquisition module, for obtaining new credit evaluation model input by user；
Matching degree determining module, it is accurate for effective information, model test coverage and the model according to the new credit evaluation model
Exactness determines the Model Matching degree of the new credit evaluation model；
First judgment module, for judging whether the Model Matching degree of the new credit evaluation model is more than former credit evaluation model
Model Matching degree, obtain the first judging result；
Processing module, for working as the model that the Model Matching degree of the new credit evaluation model is more than the former credit evaluation model
During matching degree, the former credit evaluation model is replaced with the new credit evaluation model；
When the Model Matching degree of the new credit evaluation model is less than or equal to the Model Matching of the former credit evaluation model
When spending, retain the former credit evaluation model.
6. more new system according to claim 5, which is characterized in that the matching degree computing module specifically includes：
Effective information determination unit, the Information quanlity index for each index according to the new credit evaluation model determine institute
State the effective information of new credit evaluation model；
Model test coverage determination unit, for each index according to the new credit evaluation model coverage rate determine it is described new
The model test coverage of credit evaluation model；
Model accuracy determination unit, for determining that the model of the new credit evaluation model is accurate according to each root-mean-square error
Degree, wherein, the root-mean-square error is the equal of assessment result when being assessed using the new credit evaluation model sample set
Square error；
Model Matching degree computing unit, for according to the effective information, the model test coverage and the model accuracy
Calculate the Model Matching degree of the new credit evaluation model.
7. update method according to claim 6, which is characterized in that the effective information determination unit specifically includes：
Information quanlity index determination subelement, for according to formula：Determine the new letter
With the Information quanlity index of i-th of index of assessment models, wherein, HiRepresent the Information quanlity index of i-th of index, xijRepresent i-th
Corresponding j-th of the single incident of a index, χiRepresent the corresponding event sets of i-th of index, p (xij) represent xijWhat is occurred is general
Rate；
Effective information determination subelement, for according to formula：Determine effective letter of the new credit evaluation model
Breath amount, wherein, V represents effective information, and N represents index quantity.
8. update method according to claim 6, which is characterized in that the model accuracy determination unit specifically includes：
Error determination subelement, for according to formula：Determine that each sample set corresponds to
Root-mean-square error, wherein, erriRepresent the corresponding root-mean-square error of i-th of sample set, M represents the sample in i-th of sample set
This quantity, scoreijRepresent fractional value when the new credit evaluation model assesses j-th of sample in i-th of sample set,Represent the standard score of j-th of sample in i-th of sample set；
Accuracy determination subelement, for according to formula：Determine that the model of the new credit evaluation model is accurate
Exactness, wherein, Q represents model accuracy, the quantity of N ' expression sample sets, piRepresent that the corresponding root mean square of i-th of sample set misses
The norm of difference.
9. a kind of credit estimation method based on block chain, which is characterized in that the block chain includes multiple gate systems and more
A data subchain, each gate system correspond to a data subchain, and each data subchain is for temporally suitable
Sequence records commercial matters information data, government affairs information data and the social information data of gate system corresponding with the data subchain；
The credit estimation method includes：
Obtain the credit data of gate system to be assessed, the credit data is data corresponding with the gate system to be assessed
Commercial matters information data, government affairs information data and/or the social information data recorded in subchain；
The credit evaluation value of the gate system to be assessed is determined according to the credit data and credit evaluation model, wherein, institute
It is according to the updated credit evaluation model of claim 1-4 any one of them update methods to state credit evaluation model.
10. a kind of credit evaluation system based on block chain, which is characterized in that the block chain includes multiple gate systems and more
A data subchain, each gate system correspond to a data subchain, and each data subchain is for temporally suitable
Sequence records commercial matters information data, government affairs information data and the social information data of gate system corresponding with the data subchain；
The credit evaluation system includes：
Credit data acquisition module, for obtaining the credit data of gate system to be assessed, the credit data is to be treated with described
The commercial matters information data, the government affairs information data and/or the social activity recorded in the corresponding data subchain of assessment gate system
Information data；
Evaluation module, the credit for determining the gate system to be assessed according to the credit data and credit evaluation model are commented
Valuation, wherein, the credit evaluation model is to be commented according to the updated credit of claim 1-4 any one of them update methods
Estimate model.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201711327015.9A CN108090776A (en) | 2017-12-13 | 2017-12-13 | The update method and system of credit evaluation model, credit estimation method and system |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201711327015.9A CN108090776A (en) | 2017-12-13 | 2017-12-13 | The update method and system of credit evaluation model, credit estimation method and system |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN108090776A true CN108090776A (en) | 2018-05-29 |
Family
ID=62174920
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201711327015.9A Pending CN108090776A (en) | 2017-12-13 | 2017-12-13 | The update method and system of credit evaluation model, credit estimation method and system |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN108090776A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109214883A (en) * | 2018-07-27 | 2019-01-15 | 阿里巴巴集团控股有限公司 | Service lease method, apparatus, system and electronic equipment based on block chain |
-
2017
- 2017-12-13 CN CN201711327015.9A patent/CN108090776A/en active Pending
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109214883A (en) * | 2018-07-27 | 2019-01-15 | 阿里巴巴集团控股有限公司 | Service lease method, apparatus, system and electronic equipment based on block chain |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Chen et al. | A novel grey wave forecasting method for predicting metal prices | |
| CN104657788A (en) | Method for predicting key industrial electricity consumption based on industrial condition index | |
| CN108550047A (en) | The prediction technique and device of trading volume | |
| KR101708831B1 (en) | Method for allocating a fund investment and server and computer readable recording medium using the same | |
| Baughman et al. | Deconstructing the 2017 changes to AWS spot market pricing | |
| CN103853939A (en) | Combined forecasting method for monthly load of power system based on social economic factor influence | |
| CN108256724B (en) | Power distribution network open capacity planning method based on dynamic industry coefficient | |
| CN110716998B (en) | Fine scale population data spatialization method | |
| Harkushenko et al. | Analysis of economic and mathematical models of information and communication technology (ICT) effect on the production output: Does the solow paradox exist? | |
| CN101295388A (en) | Credit estimation method and system | |
| CN111797530B (en) | Method, device, equipment and medium for determining weak links of power system | |
| CN108090776A (en) | The update method and system of credit evaluation model, credit estimation method and system | |
| Islas-Camargo et al. | Can the informal sector affect the relationship between unemployment and output? An analysis of the Mexican case | |
| Wang et al. | A Bayesian combination forecasting model for retail supply chain coordination | |
| Coates et al. | Agent-based modelling and inundation prediction to enable the identification of businesses affected by flooding | |
| CN110264056B (en) | Method and system for analyzing electric quantity consumption | |
| CN113902457B (en) | Method and device for evaluating reliability of house source information, electronic equipment and storage medium | |
| Kasperowicz et al. | Reengineering of electricity market monitoring | |
| CN108846746A (en) | A kind of carbon transaction behavior modeling method of combination discrete statistics and extreme learning machine | |
| CN110490488B (en) | Power enterprise main network planning data analysis system based on big data analysis technology | |
| CN102306353A (en) | Method and system for estimating credibility of simulation system | |
| CN108363789B (en) | Short-term missing repairing method and device for electricity consumption meter data of industrial and commercial users | |
| CN106301880A (en) | One determines that cyberrelationship degree of stability, Internet service recommend method and apparatus | |
| Thye et al. | Human capital development and income inequality in Indonesia: Evidence from a nonlinear autoregressive distributed lag (NARDL) analysis | |
| CN110837980A (en) | Enterprise credit rating method, device, equipment and storage medium |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20180529 |
|
| RJ01 | Rejection of invention patent application after publication |