CN107993140A - A kind of personal credit's methods of risk assessment and system - Google Patents
A kind of personal credit's methods of risk assessment and system Download PDFInfo
- Publication number
- CN107993140A CN107993140A CN201711171045.5A CN201711171045A CN107993140A CN 107993140 A CN107993140 A CN 107993140A CN 201711171045 A CN201711171045 A CN 201711171045A CN 107993140 A CN107993140 A CN 107993140A
- Authority
- CN
- China
- Prior art keywords
- user
- network model
- training
- generation
- risk
- 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
- G06Q40/00—Finance; Insurance; Tax strategies; Processing of corporate or income taxes
- G06Q40/03—Credit; Loans; Processing thereof
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/048—Activation functions
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Business, Economics & Management (AREA)
- General Physics & Mathematics (AREA)
- General Health & Medical Sciences (AREA)
- Accounting & Taxation (AREA)
- Software Systems (AREA)
- Mathematical Physics (AREA)
- General Engineering & Computer Science (AREA)
- Computing Systems (AREA)
- Molecular Biology (AREA)
- Finance (AREA)
- Health & Medical Sciences (AREA)
- Life Sciences & Earth Sciences (AREA)
- Artificial Intelligence (AREA)
- Biomedical Technology (AREA)
- Biophysics (AREA)
- Computational Linguistics (AREA)
- Data Mining & Analysis (AREA)
- Evolutionary Computation (AREA)
- Marketing (AREA)
- Development Economics (AREA)
- General Business, Economics & Management (AREA)
- Economics (AREA)
- Technology Law (AREA)
- Strategic Management (AREA)
- Financial Or Insurance-Related Operations Such As Payment And Settlement (AREA)
Abstract
The invention discloses a kind of personal credit's methods of risk assessment, including：Gather user's real information sample；Generation user's training data sample is simulated according to the distribution character of user's real information sample；Assessment models are trained according to user's training data sample；Obtain the real information of user to be assessed；The real information of the user to be assessed is inputted in the assessment models to training, obtains the risk assessment information of user to be assessed.The invention also discloses a kind of personal credit's risk evaluating system.The present invention can improve assessment efficiency and accuracy rate.
Description
Technical field
The present invention relates to computer and internet financial field, more particularly to a kind of personal credit's methods of risk assessment and it is
System.
Background technology
With developing rapidly for internet finance, personal credit is increasingly becoming a kind of mainstream production in internet financial field
The problem of product, network loan causes, also gradually shows, due to the risk assessment inefficiency of personal credit, the mistake artificially assessed
Rate remains high, and result in many network money-lenders can not repay the loan in time, and lending mechanism can not be obtained with Company capital
Effectively turnover.
Personal credit's risk control method of mainstream is all based on artificial appraisal procedure at present, artificial to assess single paragraph person's
Repaying ability is simultaneously inefficiency using certain fixed assets as guarantee, the method shortcoming, and there are commented caused by human error
Estimate inconsistency and erroneous judgement risk.
The content of the invention
The present invention is directed to problems of the prior art, there is provided a kind of personal credit's methods of risk assessment and system,
Assessment efficiency and accuracy rate can be improved.
The technical solution that the present invention is proposed with regard to above-mentioned technical problem is as follows：
On the one hand, the present invention provides a kind of personal credit's methods of risk assessment, including：
Gather user's real information sample；
Generation user's training data sample is simulated according to the distribution character of user's real information sample；
Assessment models are trained according to user's training data sample；
Obtain the real information of user to be assessed；
The real information of the user to be assessed is inputted in the assessment models to training, obtains the wind of user to be assessed
Danger assessment information.
Further, it is described that generation user's training data sample is simulated according to the distribution character of user's real information sample
This, specifically includes：
Establish and train generation network model；
Gaussian noise, simulation generation and user's real information sample are inputted into the generation network model after training
The consistent user's training data sample of distribution character.
Further, it is described to establish and train generation network model, specifically include：
Establish confrontation network model；The confrontation network model includes generation network model and differentiates network model；
Network weight is generated according to the generation error transfer factor of the generation network model, to minimize object function；
Network weight is differentiated according to the differentiation error transfer factor of the differentiation network model, to maximize the object function；
By being adjusted to the generation network weight and the alternating for differentiating network weight, make the generation error and institute
State and differentiate that error reaches dynamic equilibrium, complete the generation network model and the training for differentiating network model.
Further, the real information by the user to be assessed is inputted in the assessment models to training, is obtained
The risk assessment information of user to be assessed, specifically includes：
The real information of the user to be assessed is inputted in the assessment models to training, it is defeated to obtain the assessment models
The risk factor gone out；
Obtain the corresponding credit accrediting amount of the risk factor and interest rate value, and by the risk factor and corresponding letter
Borrow the risk assessment information of the accrediting amount and interest rate value as user to be assessed.
Further, after the risk assessment information for obtaining user to be assessed, further include：
The deviation data in risk assessment information to having assessed user is modified；
Training is re-started to the generation network model using revised risk assessment information, so that after re -training
Generation network model simulation generate new user's training data sample；
Periodically the assessment models are trained using new user's training data sample, are kept at the assessment models
In effective status.
On the other hand, the present invention provides a kind of personal credit's risk evaluating system, including：
Acquisition module, for gathering user's real information sample；
Generation module is simulated, number is trained for simulating generation user according to the distribution character of user's real information sample
According to sample；
Training module, for being trained according to user's training data sample to assessment models；
Acquisition module, for obtaining the real information of user to be assessed；And
Evaluation module, for inputting the real information of the user to be assessed in the assessment models to training, obtains
The risk assessment information of user to be assessed.
Further, the simulation generation module specifically includes：
Unit is established, for establishing and training generation network model；And
Generation unit, for inputting Gaussian noise, simulation generation and the user into the generation network model after training
The consistent user's training data sample of the distribution character of real information sample.
Further, the model foundation unit specifically includes：
Model foundation subelement, network model is resisted for establishing；The confrontation network model includes generation network model
With differentiation network model；
First adjustment subelement, for generating network weight according to the generation error transfer factor of the generation network model, with
Minimize object function；
Second adjustment subelement, for differentiating network weight according to the differentiation error transfer factor of the differentiation network model, with
Maximize the object function；And
Training subelement, for by being adjusted to the generation network weight and the alternating for differentiating network weight, making
The generation error and the differentiation error reach dynamic equilibrium, complete the generation network model and the differentiation network model
Training.
Further, the evaluation module specifically includes：
Risk factor acquiring unit, for the real information of the user to be assessed to be inputted the assessment models to training
In, obtain the risk factor that the assessment models export；And
Information acquisition unit is assessed, for obtaining the corresponding credit accrediting amount of the risk factor and interest rate value, and will
The risk assessment information of the risk factor and the corresponding credit accrediting amount and interest rate value as user to be assessed.
Further, personal credit's risk evaluating system further includes：
Correcting module, is modified for the deviation data in the risk assessment information to having assessed user；
Re -training module, for re-starting instruction to the generation network model using revised risk assessment information
Practice, so that the generation network model simulation after re -training generates new user's training data sample；And
Regular exercise module, for being periodically trained using new user's training data sample to the assessment models,
The assessment models are kept to be in effective status.
The beneficial effect that technical solution provided in an embodiment of the present invention is brought is：
User's real information sample is gathered, generation network model is trained, makes the simulation generation of generation network model big
The amount user training data sample consistent with the distribution character of user's real information sample, then sample is trained using substantial amounts of user
This, is trained assessment models, the assessment models after training is assessed according to the real information of user to be assessed, obtains
The risk assessment information of user to be assessed, without the artificial efficiency and accuracy assessed, effectively improve assessment.
Brief description of the drawings
To describe the technical solutions in the embodiments of the present invention more clearly, make required in being described below to embodiment
Attached drawing is briefly described, it should be apparent that, drawings in the following description are only some embodiments of the present invention, for
For those of ordinary skill in the art, without creative efforts, other can also be obtained according to these attached drawings
Attached drawing.
Fig. 1 is the flow diagram for personal credit's methods of risk assessment that the embodiment of the present invention one provides；
Fig. 2 is the schematic diagram for personal credit's methods of risk assessment that the embodiment of the present invention one provides；
Fig. 3 is the structure diagram of personal credit's risk evaluating system provided by Embodiment 2 of the present invention.
Embodiment
To make the object, technical solutions and advantages of the present invention clearer, below in conjunction with attached drawing to embodiment party of the present invention
Formula is described in further detail.
Embodiment one
An embodiment of the present invention provides a kind of personal credit's methods of risk assessment, referring to Fig. 1, this method includes：
S1, collection user's real information sample；
S2, simulate generation user's training data sample according to the distribution character of user's real information sample；
S3, according to user's training data sample be trained assessment models；
S4, the real information for obtaining user to be assessed；
S5, input the real information of the user to be assessed in the assessment models to training, obtains user to be assessed
Risk assessment information.
It should be noted that in step sl, user's real information of collection mainly includes three aspects：User itself is personal
Attribute information, social information and proprietary information.Wherein, user itself personal attribute information includes gender, age, native place, education
Background, marriage and childbirth, occupation etc.；Social information includes the information such as cell-phone number replacement frequency, address replacement frequency, hobby；Proprietary information
Situation etc. is in arrears with including house property, vehicle, loan, credit card amount, credit card.
Before step S2 is performed, structuring processing is carried out to the information gathered.For example, the text information that will be collected
Vectorization expression is carried out, all information collected carry out unified scale and represent, it is unified using day to be single to be such as in arrears with the credit card date
Position etc..And then the information of structuring is stored in database, and the user data for having missing is removed, while the number that will have redundancy
According to necessarily being deleted.
Further, in step s 2, it is described that generation use is simulated according to the distribution character of user's real information sample
Family training data sample, specifically includes：
Establish and train generation network model；
Gaussian noise, simulation generation and user's real information sample are inputted into the generation network model after training
The consistent user's training data sample of distribution character.
Further, it is described to establish and train generation network model, specifically include：
Establish confrontation network model；The confrontation network model includes generation network model and differentiates network model；
Network weight is generated according to the generation error transfer factor of the generation network model, to minimize object function；
Network weight is differentiated according to the differentiation error transfer factor of the differentiation network model, to maximize the object function；
By being adjusted to the generation network weight and the alternating for differentiating network weight, make the generation error and institute
State and differentiate that error reaches dynamic equilibrium, complete the generation network model and the training for differentiating network model.
It should be noted that anti-network model is a network model based on game, wherein, generate network model
Generato is used to generate the user instruction consistent with the distribution character of user's real information by the noise simulation of Gaussian Profile
Practice data；Differentiate that network model Discriminator is used to differentiate that the data of its input to be the information generated by Generato
Or user's real information.
Generate network model and differentiate that network model is full Connection Neural Network, structure is：
Wherein, w represents weights, and b represents biasing, and W represents the matrix of weights, the transposition of T representing matrixes, and f represents activation letter
Number, generally sigmoid functions, by multi-level stacking, form a complete neutral net.
In network model is generated, one random noise vector z with Gaussian Profile of input of neutral net, its dimension
Can be any n values, and the data dimension of user's real information is m, user's real information is expressed as { x1, x2..., xm, y },
Wherein y is the label of data, i.e. risk factor, then the output dimension for generating network model is m, and generation data format is { x1,
x2..., xm}.In network model is differentiated, input dimension is m, and output dimension is 1, that is, generates the data that network model is generated
With user's truthful data as the input for differentiating network model, output label 1 represents network and judges that input data is true as user
Data, output label represent network for 0 and judge that input data makes a living into the data of network model generation.
Generation network model and the differentiation network model object function to be optimized are：
Wherein, D represents to differentiate network model, and G represents generation network model.
When optimization generates network model, object function V is minimized, and adjust generation network weight, differentiated in optimization
During network model, object function V is maximized, and adjust differentiation network weight, by alternately updating two network weights, made
Two network models are staggeredly trained by game method, to get a promotion at the same time in game, when generation error is with differentiating error
Two network models are restrained when reaching dynamic equilibrium, are finally reached the system for the data for being generated generation network model
Cloth of scoring is extremely similar to the statistical distribution of user's truthful data.
After two network model training completions reach convergence, the generation network model in confrontation network model is extracted
Part.Random Gaussian reflectivity mirrors are inputted into generation network model, generation network model is generated several and meets assessment
The training sample of model training, i.e. user's training data sample.
Further, in step s3, assessment models are an XGBoost assessment models, and wherein XGBoost is a kind of base
In the machine learning algorithm of decision tree, feature is good etc. for that can carry out end-to-end study, parallel speed block, robustness, is one
The good classification ensemble learning algorithm of kind.In the present embodiment, be adapted to according to the feature selecting of data XGBoost parameters (by
Different with dimension in data scale, optimal parameter is also different), including iterations, learning rate, object function, maximal tree depth
Deng.
After XGBoost assessment models are established, given according to a large number of users training data sample of generation and artificial standard
The risk factor gone out is trained XGBoost assessment models.
Further, in step s 5, the real information by the user to be assessed inputs the assessment to training
In model, the risk assessment information of user to be assessed is obtained, is specifically included：
The real information of the user to be assessed is inputted in the assessment models to training, it is defeated to obtain the assessment models
The risk factor gone out；
Obtain the corresponding credit accrediting amount of the risk factor and interest rate value, and by the risk factor and corresponding letter
Borrow the risk assessment information of the accrediting amount and interest rate value as user to be assessed.
It should be noted that inputting the real information of user to be assessed into assessment models, assessment models output is to be evaluated
Estimate the risk factor of user, the final credit information of user to be assessed obtained according to risk factor, such as maximum accrediting amount and
Interest rate relation etc..
Further, after step s 5, i.e., also wrapped after the risk assessment information for obtaining user to be assessed
Include：
The deviation data in risk assessment information to having assessed user is modified；
Training is re-started to the generation network model using revised risk assessment information, so that after re -training
Generation network model simulation generate new user's training data sample；
Periodically the assessment models are trained using new user's training data sample, are kept at the assessment models
In effective status.
It should be noted that after assessment is made to user to be assessed, assessment result can be also modified, and according to repairing
Result re -training generation network model after just, assessment models, further improve the accuracy of assessment.
Specifically, after assessment, the risk assessment information for having assessed user is all preserved in the database, and periodic cleaning
Partial failure historical data.Artificially the deviation data in risk assessment information is modified, such as picks out risk assessment letter
There are risk factor and data of obvious deviation etc. in breath.The training of generation network model is re-started according to revised data,
It is valid data always to keep training data, and then periodically continues trained XGBoost assessment models using new training data, is protected
Hold assessment models and be in long-time effective status.
Referring to Fig. 2, for the schematic diagram of personal credit's methods of risk assessment provided in an embodiment of the present invention.Generate network model
Gaussian noise is inputted, differentiates that network model input generates data and the user's truthful data that network model is generated, to generation
Network model and differentiation network model are alternately trained, and are made two network models in game while are improved, realize unsupervised
Study.After the training of generation network model is completed, Gaussian noise is inputted into generation network model, makes generation network model defeated
Go out training data largely consistent with the distribution character of user's truthful data.Further according to training data to XGBoost assessment models
It is trained, user information is inputted to the XGBoost assessment models to training, you can export by XGBoost assessment models
Assessment result, obtains the risk assessment information of user.
The embodiment of the present invention gathers user's real information sample, and generation network model is trained, makes generation network mould
Pattern is intended generating user's training data sample largely consistent with the distribution character of user's real information sample, then using a large amount of
User's training sample, assessment models are trained, make the assessment models after training according to the real information of user to be assessed
Assessed, obtain the risk assessment information of user to be assessed, without artificial assessment, effectively improve the efficiency of assessment and accurate
Property.
Embodiment two
An embodiment of the present invention provides a kind of personal credit's risk evaluating system, can realize that above-mentioned personal credit's risk is commented
Estimate all flows of method, referring to Fig. 3, the system comprises：
Acquisition module 1, for gathering user's real information sample；
Generation module 2 is simulated, for simulating generation user's training according to the distribution character of user's real information sample
Data sample；
Training module 3, for being trained according to user's training data sample to assessment models；
Acquisition module 4, for obtaining the real information of user to be assessed；And
Evaluation module 5, for inputting the real information of the user to be assessed in the assessment models to training, obtains
The risk assessment information of user to be assessed.
Further, the simulation generation module specifically includes：
Unit is established, for establishing and training generation network model；And
Generation unit, for inputting Gaussian noise, simulation generation and the user into the generation network model after training
The consistent user's training data sample of the distribution character of real information sample.
Further, the model foundation unit specifically includes：
Model foundation subelement, network model is resisted for establishing；The confrontation network model includes generation network model
With differentiation network model；
First adjustment subelement, for generating network weight according to the generation error transfer factor of the generation network model, with
Minimize object function；
Second adjustment subelement, for differentiating network weight according to the differentiation error transfer factor of the differentiation network model, with
Maximize the object function；And
Training subelement, for by being adjusted to the generation network weight and the alternating for differentiating network weight, making
The generation error and the differentiation error reach dynamic equilibrium, complete the generation network model and the differentiation network model
Training.
Further, the evaluation module specifically includes：
Risk factor acquiring unit, for the real information of the user to be assessed to be inputted the assessment models to training
In, obtain the risk factor that the assessment models export；And
Information acquisition unit is assessed, for obtaining the corresponding credit accrediting amount of the risk factor and interest rate value, and will
The risk assessment information of the risk factor and the corresponding credit accrediting amount and interest rate value as user to be assessed.
Further, personal credit's risk evaluating system further includes：
Correcting module, is modified for the deviation data in the risk assessment information to having assessed user；
Re -training module, for re-starting instruction to the generation network model using revised risk assessment information
Practice, so that the generation network model simulation after re -training generates new user's training data sample；And
Regular exercise module, for being periodically trained using new user's training data sample to the assessment models,
The assessment models are kept to be in effective status.
The embodiment of the present invention gathers user's real information sample, and generation network model is trained, makes generation network mould
Pattern is intended generating user's training data sample largely consistent with the distribution character of user's real information sample, then using a large amount of
User's training sample, assessment models are trained, make the assessment models after training according to the real information of user to be assessed
Assessed, obtain the risk assessment information of user to be assessed, without artificial assessment, effectively improve the efficiency of assessment and accurate
Property.
The foregoing is merely presently preferred embodiments of the present invention, is not intended to limit the invention, it is all the present invention spirit and
Within principle, any modification, equivalent replacement, improvement and so on, should all be included in the protection scope of the present invention.
Claims (10)
- A kind of 1. personal credit's methods of risk assessment, it is characterised in that including：Gather user's real information sample；Generation user's training data sample is simulated according to the distribution character of user's real information sample；Assessment models are trained according to user's training data sample；Obtain the real information of user to be assessed；The real information of the user to be assessed is inputted in the assessment models to training, the risk for obtaining user to be assessed is commented Estimate information.
- 2. personal credit's methods of risk assessment as claimed in claim 1, it is characterised in that described truly to be believed according to the user Distribution character simulation generation user's training data sample of sample is ceased, is specifically included：Establish and train generation network model；Gaussian noise, simulation generation and the distribution of user's real information sample are inputted into the generation network model after training The consistent user's training data sample of characteristic.
- 3. personal credit's methods of risk assessment as claimed in claim 2, it is characterised in that described to establish and train generation network Model, specifically includes：Establish confrontation network model；The confrontation network model includes generation network model and differentiates network model；Network weight is generated according to the generation error transfer factor of the generation network model, to minimize object function；Network weight is differentiated according to the differentiation error transfer factor of the differentiation network model, to maximize the object function；By being adjusted to the generation network weight and the alternating for differentiating network weight, make the generation error and described sentence Other error reaches dynamic equilibrium, completes the generation network model and the training for differentiating network model.
- 4. personal credit's methods of risk assessment as claimed in claim 1, it is characterised in that described by the user's to be assessed Real information is inputted in the assessment models to training, is obtained the risk assessment information of user to be assessed, is specifically included：The real information of the user to be assessed is inputted in the assessment models to training, obtains the assessment models output Risk factor；The corresponding credit accrediting amount of the risk factor and interest rate value are obtained, and the risk factor and corresponding credit are awarded Believe the risk assessment information of amount and interest rate value as user to be assessed.
- 5. personal credit's methods of risk assessment as claimed in claim 2, it is characterised in that obtain user's to be assessed described After risk assessment information, further include：The deviation data in risk assessment information to having assessed user is modified；Training is re-started to the generation network model using revised risk assessment information, so that the life after re -training New user's training data sample is generated into network modeling；Periodically the assessment models are trained using new user's training data sample, keeping the assessment models to be in has Effect state.
- A kind of 6. personal credit's risk evaluating system, it is characterised in that including：Acquisition module, for gathering user's real information sample；Generation module is simulated, for simulating generation user's training data sample according to the distribution character of user's real information sample This；Training module, for being trained according to user's training data sample to assessment models；Acquisition module, for obtaining the real information of user to be assessed；AndEvaluation module, for inputting the real information of the user to be assessed in the assessment models to training, obtains to be evaluated Estimate the risk assessment information of user.
- 7. personal credit's risk evaluating system as claimed in claim 6, it is characterised in that the simulation generation module specifically wraps Include：Unit is established, for establishing and training generation network model；AndGeneration unit, for inputting Gaussian noise into the generation network model after training, simulation generation is true with the user The consistent user's training data sample of the distribution character of message sample.
- 8. personal credit's risk evaluating system as claimed in claim 7, it is characterised in that the model foundation unit specifically wraps Include：Model foundation subelement, network model is resisted for establishing；The confrontation network model includes generation network model and sentences Other network model；First adjustment subelement, for generating network weight according to the generation error transfer factor of the generation network model, with minimum Change object function；Second adjustment subelement, for differentiating network weight according to the differentiation error transfer factor of the differentiation network model, with maximum Change the object function；AndTraining subelement, for by being adjusted to the generation network weight and the alternating for differentiating network weight, making described Generation error and the differentiation error reach dynamic equilibrium, complete the generation network model and the instruction for differentiating network model Practice.
- 9. personal credit's risk evaluating system as claimed in claim 6, it is characterised in that the evaluation module specifically includes：Risk factor acquiring unit, for inputting the real information of the user to be assessed in the assessment models to training, Obtain the risk factor of the assessment models output；AndInformation acquisition unit is assessed, for obtaining the corresponding credit accrediting amount of the risk factor and interest rate value, and by described in The risk assessment information of risk factor and the corresponding credit accrediting amount and interest rate value as user to be assessed.
- 10. personal credit's risk evaluating system as claimed in claim 7, it is characterised in that personal credit's risk assessment System further includes：Correcting module, is modified for the deviation data in the risk assessment information to having assessed user；Re -training module, for re-starting training to the generation network model using revised risk assessment information, So that the generation network model simulation after re -training generates new user's training data sample；AndRegular exercise module, for being periodically trained using new user's training data sample to the assessment models, is kept The assessment models are in effective status.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201711171045.5A CN107993140A (en) | 2017-11-22 | 2017-11-22 | A kind of personal credit's methods of risk assessment and system |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201711171045.5A CN107993140A (en) | 2017-11-22 | 2017-11-22 | A kind of personal credit's methods of risk assessment and system |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN107993140A true CN107993140A (en) | 2018-05-04 |
Family
ID=62032520
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201711171045.5A Pending CN107993140A (en) | 2017-11-22 | 2017-11-22 | A kind of personal credit's methods of risk assessment and system |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN107993140A (en) |
Cited By (13)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108648068A (en) * | 2018-05-16 | 2018-10-12 | 长沙农村商业银行股份有限公司 | A kind of assessing credit risks method and system |
| CN108765340A (en) * | 2018-05-29 | 2018-11-06 | Oppo（重庆）智能科技有限公司 | Fuzzy image processing method, apparatus and terminal device |
| CN109255506A (en) * | 2018-11-22 | 2019-01-22 | 重庆邮电大学 | A kind of internet finance user's overdue loan prediction technique based on big data |
| CN109919196A (en) * | 2019-02-01 | 2019-06-21 | 华南理工大学 | A kind of constitution recognition methods based on feature selecting and disaggregated model |
| CN110135972A (en) * | 2019-04-23 | 2019-08-16 | 上海淇玥信息技术有限公司 | A kind of method, apparatus, system and recording medium for improving user and moving branch rate |
| CN110310199A (en) * | 2019-06-27 | 2019-10-08 | 上海上湖信息技术有限公司 | Borrow or lend money construction method, system and the debt-credit Risk Forecast Method of risk forecast model |
| CN110322055A (en) * | 2019-06-18 | 2019-10-11 | 阿里巴巴集团控股有限公司 | A kind of method and system improving data risk model scoring stability |
| WO2019232892A1 (en) * | 2018-06-05 | 2019-12-12 | 平安科技（深圳）有限公司 | Method and device for estimating risk probability associated with insurance purchaser, computer apparatus, and storage medium |
| CN110633989A (en) * | 2019-08-16 | 2019-12-31 | 阿里巴巴集团控股有限公司 | Method and device for determining risk behavior generation model |
| CN111126503A (en) * | 2019-12-27 | 2020-05-08 | 北京同邦卓益科技有限公司 | Training sample generation method and device |
| CN111429270A (en) * | 2020-04-22 | 2020-07-17 | 广州东百信息科技有限公司 | Overseas credit card wind control model acquisition method, device, equipment and storage medium |
| CN112308203A (en) * | 2019-10-09 | 2021-02-02 | 刘畅 | Bank loan issuing and post-loan management decision support system based on artificial intelligence deep learning and multi-parameter dynamic game |
| CN112347343A (en) * | 2020-09-25 | 2021-02-09 | 北京淇瑀信息科技有限公司 | Customized information pushing method and device and electronic equipment |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US9262688B1 (en) * | 2011-09-24 | 2016-02-16 | Z Advanced Computing, Inc. | Method and system for analyzing and recognition of an emotion or expression from multimedia, text, or sound track |
| CN106127363A (en) * | 2016-06-12 | 2016-11-16 | 腾讯科技（深圳）有限公司 | A kind of user credit appraisal procedure and device |
| CN106651574A (en) * | 2016-12-30 | 2017-05-10 | 苏州大学 | Personal credit assessment method and apparatus |
| CN107025598A (en) * | 2017-04-06 | 2017-08-08 | 中国矿业大学 | A kind of individual credit risk appraisal procedure based on extreme learning machine |
| CN107292528A (en) * | 2017-06-30 | 2017-10-24 | 阿里巴巴集团控股有限公司 | Vehicle insurance Risk Forecast Method, device and server |
-
2017
- 2017-11-22 CN CN201711171045.5A patent/CN107993140A/en active Pending
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US9262688B1 (en) * | 2011-09-24 | 2016-02-16 | Z Advanced Computing, Inc. | Method and system for analyzing and recognition of an emotion or expression from multimedia, text, or sound track |
| CN106127363A (en) * | 2016-06-12 | 2016-11-16 | 腾讯科技（深圳）有限公司 | A kind of user credit appraisal procedure and device |
| CN106651574A (en) * | 2016-12-30 | 2017-05-10 | 苏州大学 | Personal credit assessment method and apparatus |
| CN107025598A (en) * | 2017-04-06 | 2017-08-08 | 中国矿业大学 | A kind of individual credit risk appraisal procedure based on extreme learning machine |
| CN107292528A (en) * | 2017-06-30 | 2017-10-24 | 阿里巴巴集团控股有限公司 | Vehicle insurance Risk Forecast Method, device and server |
Non-Patent Citations (1)
| Title |
|---|
| 王坤峰 等: ""生成式对抗网络GAN的研究进展与展望"", 《自动化学报》 * |
Cited By (19)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108648068A (en) * | 2018-05-16 | 2018-10-12 | 长沙农村商业银行股份有限公司 | A kind of assessing credit risks method and system |
| CN108765340A (en) * | 2018-05-29 | 2018-11-06 | Oppo（重庆）智能科技有限公司 | Fuzzy image processing method, apparatus and terminal device |
| WO2019232892A1 (en) * | 2018-06-05 | 2019-12-12 | 平安科技（深圳）有限公司 | Method and device for estimating risk probability associated with insurance purchaser, computer apparatus, and storage medium |
| CN109255506A (en) * | 2018-11-22 | 2019-01-22 | 重庆邮电大学 | A kind of internet finance user's overdue loan prediction technique based on big data |
| CN109919196A (en) * | 2019-02-01 | 2019-06-21 | 华南理工大学 | A kind of constitution recognition methods based on feature selecting and disaggregated model |
| CN109919196B (en) * | 2019-02-01 | 2023-12-08 | 华南理工大学 | Physique identification method based on feature selection and classification model |
| CN110135972A (en) * | 2019-04-23 | 2019-08-16 | 上海淇玥信息技术有限公司 | A kind of method, apparatus, system and recording medium for improving user and moving branch rate |
| CN110322055A (en) * | 2019-06-18 | 2019-10-11 | 阿里巴巴集团控股有限公司 | A kind of method and system improving data risk model scoring stability |
| CN110322055B (en) * | 2019-06-18 | 2024-02-09 | 创新先进技术有限公司 | Method and system for improving grading stability of data risk model |
| CN110310199B (en) * | 2019-06-27 | 2022-03-18 | 上海上湖信息技术有限公司 | Method and system for constructing loan risk prediction model and loan risk prediction method |
| CN110310199A (en) * | 2019-06-27 | 2019-10-08 | 上海上湖信息技术有限公司 | Borrow or lend money construction method, system and the debt-credit Risk Forecast Method of risk forecast model |
| CN110633989A (en) * | 2019-08-16 | 2019-12-31 | 阿里巴巴集团控股有限公司 | Method and device for determining risk behavior generation model |
| CN110633989B (en) * | 2019-08-16 | 2023-08-29 | 创新先进技术有限公司 | Determination method and device for risk behavior generation model |
| CN112308203A (en) * | 2019-10-09 | 2021-02-02 | 刘畅 | Bank loan issuing and post-loan management decision support system based on artificial intelligence deep learning and multi-parameter dynamic game |
| CN111126503B (en) * | 2019-12-27 | 2023-09-26 | 北京同邦卓益科技有限公司 | Training sample generation method and device |
| CN111126503A (en) * | 2019-12-27 | 2020-05-08 | 北京同邦卓益科技有限公司 | Training sample generation method and device |
| CN111429270A (en) * | 2020-04-22 | 2020-07-17 | 广州东百信息科技有限公司 | Overseas credit card wind control model acquisition method, device, equipment and storage medium |
| CN112347343A (en) * | 2020-09-25 | 2021-02-09 | 北京淇瑀信息科技有限公司 | Customized information pushing method and device and electronic equipment |
| CN112347343B (en) * | 2020-09-25 | 2024-05-28 | 北京淇瑀信息科技有限公司 | Custom information pushing method and device and electronic equipment |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN107993140A (en) | A kind of personal credit's methods of risk assessment and system | |
| Wang et al. | A Comparative Assessment of Credit Risk Model Based on Machine Learning——a case study of bank loan data | |
| CN108229268A (en) | Expression Recognition and convolutional neural networks model training method, device and electronic equipment | |
| CN110188794B (en) | Deep learning model training method, device, equipment and storage medium | |
| CN110276679A (en) | A kind of network individual credit fraud detection method towards deep learning | |
| Min et al. | DeepStealth: Leveraging deep learning models for stealth assessment in game-based learning environments | |
| CN105046366B (en) | model training method and device | |
| CN108596420A (en) | A kind of talent assessment system and method for Behavior-based control | |
| CN109766424A (en) | Filtering method and device for reading understanding model training data | |
| Arora et al. | PREDICTION & WARNING: a method to improve student's performance | |
| Su | Analysis of optimisation method for online education data mining based on big data assessment technology | |
| CN108765127A (en) | A kind of credit scoring card feature selection approach based on monte-carlo search | |
| Rong | Distance learning quality assessment of universities based on interval monotonic decision tree algorithm | |
| CN113409157B (en) | Cross-social network user alignment method and device | |
| Bataev et al. | Artificial intelligence technologies in higher education institutions: a model of adaptive education | |
| CN116306543B (en) | Form data generation method and system based on generation type countermeasure network | |
| CN117787715A (en) | Financial wind control method and system for generating customer portraits based on information framework | |
| Chen et al. | Learning from designers: Fashion compatibility analysis via dataset distillation | |
| Zhao | Research on Construction of Educational Management Model Based on Data Mining Technology | |
| Zhou et al. | Human motion variation synthesis with multivariate Gaussian processes | |
| Wang et al. | [Retracted] Font Design in Visual Communication Design of Genetic Algorithm | |
| CN113256404A (en) | Data processing method and device | |
| Yangxiaoxiao | Image modal analysis in art design and image recognition using AI techniques | |
| CN115830400B (en) | Data identification method and system based on federal learning mechanism | |
| Arkhipova | Information Security Specialist Readiness Indicator as a Part of the Trusted Digital Environment |
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 | ||
| CB02 | Change of applicant information | ||
| CB02 | Change of applicant information |
Address after: 1202, 12 / F, Hangsheng science and technology building, No. 8, Gaoxin South Sixth Road, high tech Zone community, Yuehai street, Nanshan District, Shenzhen City, Guangdong Province Applicant after: Shenzhen yunnet bate Technology Co.,Ltd. Address before: 518000 Guangdong city of Shenzhen province Qianhai Shenzhen Hong Kong cooperation zone before Bay Road No. 1 building 201 room A (located in Shenzhen Qianhai business secretary Co. Ltd.) Applicant before: SHENZHEN NAIFEI TECHNOLOGY Co.,Ltd. |
|
| RJ01 | Rejection of invention patent application after publication | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20180504 |