CN110322335A - A kind of credit customer qualification classification method passing through machine learning based on WOE conversion - Google Patents
A kind of credit customer qualification classification method passing through machine learning based on WOE conversion Download PDFInfo
- Publication number
- CN110322335A CN110322335A CN201910297368.1A CN201910297368A CN110322335A CN 110322335 A CN110322335 A CN 110322335A CN 201910297368 A CN201910297368 A CN 201910297368A CN 110322335 A CN110322335 A CN 110322335A
- Authority
- CN
- China
- Prior art keywords
- data
- client
- qualification
- module
- model
- 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
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
- G06F18/214—Generating training patterns; Bootstrap methods, e.g. bagging or boosting
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
-
- 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
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Data Mining & Analysis (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Business, Economics & Management (AREA)
- General Engineering & Computer Science (AREA)
- Evolutionary Biology (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Evolutionary Computation (AREA)
- Bioinformatics & Computational Biology (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Finance (AREA)
- Accounting & Taxation (AREA)
- Life Sciences & Earth Sciences (AREA)
- Artificial Intelligence (AREA)
- Economics (AREA)
- General Business, Economics & Management (AREA)
- Technology Law (AREA)
- Strategic Management (AREA)
- Marketing (AREA)
- Development Economics (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention discloses a kind of credit customer qualification classification methods for passing through machine learning based on WOE conversion, it include: data preparation and preprocessing module, model training and evaluation and test module, model deployment module, into part data processing module, client qualification division module, request for data, collage-credit data and message registration are first calculated initial data one by the data preparation and preprocessing module, initial data two is calculated by client's classification and refund data, initial data one and initial data two are subjected to data prediction, the present invention relates to qualification sorting technique fields.This passes through the credit customer qualification classification method of machine learning based on WOE conversion, it provides and is based on machine learning method, the system for realizing different qualification client segmentations, the workload of manual examination and verification can be reduced, examination & approval efficiency is improved, is learnt in time according to newly-increased customer information, the variation of client qualification is realized adaptive, manual examination and verification efficiency can be improved to a greater extent, reduce cost of labor.
Description
Technical field
The present invention relates to qualification sorting technique field, specially a kind of credit visitor that machine learning is passed through based on WOE conversion
Family qualification classification method.
Background technique
With the development of credit industry, there are more and more loan applications for lending mechanism.Traditional checking method is
Manual examination and verification are combined with scorecard, conventional method low efficiency, not sensitive enough for customer data variation.Therefore one kind is just needed
Learnt automatically according to client's situation of change, the system of indirect labor's audit improves examination & approval efficiency, optimize approval process.In addition may be used
Excavating customer information with depth has certain help for expanding objective group.
Summary of the invention
(1) the technical issues of solving
In view of the deficiencies of the prior art, the present invention provides a kind of credit customers for passing through machine learning based on WOE conversion
Qualification classification method solves the problem of manual examination and verification low efficiency, high labor cost.
(2) technical solution
In order to achieve the above object, the present invention is achieved by the following technical programs: one kind passing through machine based on WOE conversion
The credit customer qualification classification method of study, comprising:
Data preparation and preprocessing module: including request for data, collage-credit data, message registration and refund data, pass through number
According to cleaning with segmentation and WOE conversion to reach data normalization;
Model training and evaluation and test module: including lasso feature selecting, logistic regression, immediately forest, XGBoost, depth
Module and model evaluating module are practised, Model Parameter Optimization is added averaging with four kinds of prediction results that training function obtains
Value, compares evaluation and test with the practical qualification of client, obtains the indexs such as the ROC, accurate rate, recall rate of conjunctive model, and pass through this
A little indexs carry out model selection, find out optimal models;
Model deployment module: it scores and shows into part system docking and client including model loading module, client, by what is found out
Optimal models are deployed in server using Django frame, the function which mainly includes have the load of trained model, with
Client shows into part system docking, client qualification scoring;
Into part data processing module: for newly into part customer data, first being pre-processed data, data sectional, WOE
Conversion and normalized, rule and data preparation and preprocessing module rule;
Client qualification division module: the prediction result scoring obtained according to model training and evaluation and test module is divided into ten sections,
The client qualification scoring obtained according to model deployment module, the lending amount of money and manual examination and verification suggestion by setting, obtain newly into
Should the make loans amount of money and the manual examination and verification suggestion of part client.
Preferably, the data preparation first calculates request for data, collage-credit data and message registration with preprocessing module
Initial data one calculates initial data two by client's classification and refund data, by initial data one and initial data two into
Line number Data preprocess, to abnormal data elimination and similar categorization data merge, by continuous data according to reasonable segmentation rule
Then data are segmented, when more for data, numerical value value range is very big, random noise can be generated, to data sectional
Noise can be eliminated.
Preferably, the data preparation and preprocessing module split data into five parts, randomly choose a part as survey
Data are tried, remaining four part carries out WOE conversion and normalized as training data, for training data, and according to training
The calculated WOE transformation rule of the WOE of data acts on test data, similarly by the normalization acting rules of training data in survey
Try data, the combination of corresponding common property raw five kinds different training datas and test data.
Preferably, data a copy of it that the model training and evaluation and test module generate data preparation and preprocessing module
The module is inputted, feature selecting is carried out by the lasso feature selecting function in the module, selects and classifies for client qualification
Useful feature carries out next step model training.
Preferably, five parts of data that data preparation and preprocessing module generate are carried out according to feature selecting result respectively special
Model Parameter Optimization in the module is inputted after sign selection and training function carries out model training, and wherein model is different by four kinds
It is respectively logistic regression, random forest, XGBoost and deep learning that model, which is constituted,.
Preferably, the model deployment module sends client into part customer data into part system, will be into part customer data
By obtaining processed customer data into part data processing module, by the model of this partial data input model load function
In, finally obtain client qualification scoring.
Preferably, the client qualification scoring that the model deployment module goes out in showing function according to model prediction passes through visitor
Family qualification division module obtains the qualification grade of client, the lending amount of money, manual examination and verification suggestion.
Preferably, the client qualification division module calculates the accounting of the fine or not qualification client of different segmentations, with reference to history
The lending amount of money, the revenue amount of average customer calculate the investment return ratio of different segmentations from the loss amount of money, according to investment return ratio
And different grades of credit standard is assigned with reference to client of the credit product policy for different grades, for the visitor of different segmentations
Family sets different the lending amount of money and manual examination and verification suggestion.
(3) beneficial effect
The present invention provides a kind of credit customer qualification classification methods for passing through machine learning based on WOE conversion.Have with
It is lower the utility model has the advantages that
(1), the credit customer qualification classification method that machine learning should be passed through based on WOE conversion, with reference to history average customer
The lending amount of money, revenue amount from the loss amount of money calculate the investment return ratios of different segmentations, according to investment return when with reference to believing
It borrows product policy and assigns different grades of credit standard for the client of different grades, the client of different segmentations is set not
The same lending amount of money and manual examination and verification suggestion, can be effectively reduced manual examination and verification cost, improve audit accuracy rate to a certain extent,
And then credit risk is reduced, realize scientific management customer risk.
(2), the credit customer qualification classification method of machine learning, data prediction link pair should be passed through based on WOE conversion
WOE conversion has been carried out in data, has reduced influence of noise, and dimension is converted for nonumeric type data conversion comparison ONE_HOT
Less, the influence compared to conventional machines study different characteristic for client qualification final classification is poor explanatory, and WOE is converted
The influence degree that character pair classifies for client qualification can be intuitively found out afterwards, be more convenient for understanding and be also more convenient for combining artificial warp
Test optimization machine learning algorithm.
(3), the credit customer qualification classification method that should pass through machine learning based on WOE conversion, uses machine learning algorithm
Comparison tradition scoring mode card can be realized model and learn automatically, and more sensitive for customer data variation, predictablity rate is higher,
It is more acurrate for the client segmentation of different qualifications, manual examination and verification efficiency can be improved to a greater extent, reduce cost of labor.
Detailed description of the invention
Fig. 1 is invention software module rack composition.
Specific embodiment
Following will be combined with the drawings in the embodiments of the present invention, and technical solution in the embodiment of the present invention carries out clear, complete
Site preparation description, it is clear that described embodiments are only a part of the embodiments of the present invention, instead of all the embodiments.It is based on
Embodiment in the present invention, it is obtained by those of ordinary skill in the art without making creative efforts every other
Embodiment shall fall within the protection scope of the present invention.
Referring to Fig. 1, the present invention provides a kind of technical solution: a kind of credit visitor passing through machine learning based on WOE conversion
Family qualification classification method, comprising:
Data preparation and preprocessing module: including request for data, collage-credit data, message registration and refund data, pass through number
According to cleaning with segmentation and WOE conversion to reach data normalization;
Model training and evaluation and test module: including lasso feature selecting, logistic regression, immediately forest, XGBoost, depth
Module and model evaluating module are practised, Model Parameter Optimization is added averaging with four kinds of prediction results that training function obtains
Value, compares evaluation and test with the practical qualification of client, obtains the indexs such as the ROC, accurate rate, recall rate of conjunctive model, and pass through this
A little indexs carry out model selection, find out optimal models;
Model deployment module: it scores and shows into part system docking and client including model loading module, client, by what is found out
Optimal models are deployed in server using Django frame, the function which mainly includes have the load of trained model, with
Client shows into part system docking, client qualification scoring;
Into part data processing module: for newly into part customer data, first being pre-processed data, data sectional, WOE
Conversion and normalized, rule and data preparation and preprocessing module rule;
Client qualification division module: the prediction result scoring obtained according to model training and evaluation and test module is divided into ten sections,
The client qualification scoring obtained according to model deployment module, the lending amount of money and manual examination and verification suggestion by setting, obtain newly into
Should the make loans amount of money and the manual examination and verification suggestion of part client.
Request for data, collage-credit data and message registration are first calculated initial data one by data preparation and preprocessing module,
Initial data two is calculated by client's classification and refund data, initial data one and initial data two are subjected to data and located in advance
Reason, to abnormal data elimination and similar categorization data merge, by continuous data according to reasonable chopping rule to data into
Row segmentation, when more for data, numerical value value range is very big, can generate random noise, can eliminate and make an uproar to data sectional
Sound.
Data preparation and preprocessing module split data into five parts, and random selection a part is used as test data, remaining
Four parts carry out WOE conversion and normalized as training data, for training data, and are counted according to the WOE of training data
The WOE transformation rule of calculating acts on test data, similarly by the normalization acting rules of training data in test data, accordingly
Raw five kinds of common property different training datas and test data combination.
Data a copy of it that data preparation and preprocessing module generate is inputted the module by model training and evaluation and test module,
Feature selecting is carried out by the lasso feature selecting function in the module, is selected for client qualification classification useful feature
Carry out next step model training.
Five parts of data that data preparation and preprocessing module generate are subjected to feature selecting according to feature selecting result respectively
The Model Parameter Optimization and training function inputted in the module afterwards carries out model training, and wherein model is by four kinds of different model structures
At respectively logistic regression, random forest, XGBoost and deep learning.
Model deployment module sends client into part customer data into part system, will be into part customer data by into number of packages evidence
Processing module obtains processed customer data, by the model of this partial data input model load function, finally obtains visitor
The scoring of family qualification.
The client qualification scoring that model deployment module goes out in showing function according to model prediction is divided by client qualification
Module obtains the qualification grade of client, the lending amount of money, manual examination and verification suggestion.
Client qualification division module calculates the accounting of the fine or not qualification client of different segmentations, with reference to putting for history average customer
Monetary allowance volume, revenue amount calculate the investment return ratio of different segmentations from the loss amount of money, and credit production is when referred to according to investment return
Product policy assigns different grades of credit standard for the client of different grades, the client of different segmentations is set different
The lending amount of money and manual examination and verification suggestion.
In use, data elder generation input data is prepared to carry out original data processing analysis with preprocessing module, by that will count
According to dividing quinquepartite, random selection a part is used as test data, remaining four part as training data, for training data into
The data of row WOE conversion and normalized, and test is acted on according to the calculated WOE transformation rule of the WOE of training data
Data, similarly by the normalization acting rules of training data in test data, raw five kinds different training datas of corresponding common property with
The combination of test data carries out feature selecting by the lasso feature selecting function in the module, selects for client qualification
Useful feature of classifying carries out next step model training, the data input model training that will be obtained in model training and evaluation and test module
In evaluation and test module, Model Parameter Optimization and training function in the module are inputted after carrying out feature selecting according to feature selecting result
It can be carried out model training, it will be into part customer data by obtaining processed customer data into part data processing module, by this portion
Divided data input model loads in the model of function, finally obtains client qualification scoring, the client qualification gone out according to model prediction
Scoring obtains qualification grade, the lending amount of money, manual examination and verification suggestion of client by client qualification division module, and by mold portion
Administration's processed data of inside modules are inputted to be rearranged into part data processing module, using rule with data preparation and in advance
Data after handling same rule process input client qualification division module, the prediction that model training and evaluation and test module are obtained
As a result scoring is divided into ten sections, calculates the accounting of the fine or not qualification client of different segmentations, with reference to the lending gold of history average customer
Volume, revenue amount calculate the investment return ratio of different segmentations from the loss amount of money, and credit product political affairs are when referred to according to investment return
Plan assigns different grades of credit standard for the client of different grades, and different lendings is set for the client of different segmentations
The amount of money and manual examination and verification suggestion, and the client qualification scoring that the later period is obtained using model deployment module, pass through the lending of setting
The amount of money and manual examination and verification suggestion obtain should the make loans amount of money and the manual examination and verification suggestion newly into part client.
It should be noted that, in this document, relational terms such as first and second and the like are used merely to a reality
Body or operation are distinguished with another entity or operation, are deposited without necessarily requiring or implying between these entities or operation
In any actual relationship or order or sequence.Moreover, the terms "include", "comprise" or its any other variant are intended to
Non-exclusive inclusion, so that the process, method, article or equipment including a series of elements is not only wanted including those
Element, but also including other elements that are not explicitly listed, or further include for this process, method, article or equipment
Intrinsic element.In the absence of more restrictions.By sentence " element limited including one ..., it is not excluded that
There is also other identical elements in the process, method, article or apparatus that includes the element ".
It although an embodiment of the present invention has been shown and described, for the ordinary skill in the art, can be with
A variety of variations, modification, replacement can be carried out to these embodiments without departing from the principles and spirit of the present invention by understanding
And modification, the scope of the present invention is defined by the appended.
Claims (8)
1. a kind of credit customer qualification classification method for passing through machine learning based on WOE conversion, it is characterised in that: include:
Data preparation and preprocessing module: clear by data including request for data, collage-credit data, message registration and refund data
It washes with segmentation and WOE conversion to reach data normalization;
Model training and evaluation and test module: including lasso feature selecting, logistic regression, immediately forest, XGBoost, deep learning mould
Model Parameter Optimization is added with four kinds of prediction results that training function obtains and averages by block and model evaluating module, with
The practical qualification of client compares evaluation and test, obtains the indexs such as the ROC, accurate rate, recall rate of conjunctive model, and pass through these indexs
Model selection is carried out, optimal models are found out;
Model deployment module: scoring into part system docking and client and show including model loading module, client, optimal by what is found out
Model is deployed in server using Django frame, and the function which mainly includes has trained model load and client
It is shown into part system docking, client qualification scoring;
Into part data processing module: for newly into part customer data, first being pre-processed data, data sectional, WOE conversion
With normalized, rule and data preparation and preprocessing module rule;
Client qualification division module: the prediction result scoring obtained according to model training and evaluation and test module is divided into ten sections, according to
The client qualification scoring that model deployment module obtains, the lending amount of money and manual examination and verification suggestion by setting obtain newly into part visitor
Should the make loans amount of money and the manual examination and verification suggestion at family.
2. a kind of credit customer qualification classification method for passing through machine learning based on WOE conversion according to claim 1,
Be characterized in that: request for data, collage-credit data and message registration are first calculated original number by the data preparation and preprocessing module
According to one, initial data two is calculated by client's classification and refund data, initial data one and initial data two are subjected to data
Pretreatment, to abnormal data elimination and similar categorization data merge, by continuous data according to reasonable chopping rule logarithm
According to being segmented, when more for data, numerical value value range is very big, can generate random noise, can disappear to data sectional
Except noise.
3. a kind of credit customer qualification classification method for passing through machine learning based on WOE conversion according to claim 2,
Be characterized in that: the data preparation and preprocessing module split data into five parts, and random selection a part is used as test data,
Remaining four part carries out WOE conversion and normalized as training data, for training data, and according to training data
The calculated WOE transformation rule of WOE acts on test data, similarly by the normalization acting rules of training data in test number
According to the combination of corresponding common property raw five kinds different training datas and test data.
4. a kind of credit customer qualification classification method for passing through machine learning based on WOE conversion according to claim 3,
Be characterized in that: data a copy of it input that data preparation and preprocessing module generate is somebody's turn to do by the model training and evaluation and test module
Module carries out feature selecting by lasso feature selecting function in the module, select classify for client qualification it is useful
Feature carries out next step model training.
5. a kind of credit customer qualification classification method for passing through machine learning based on WOE conversion according to claim 4,
It is characterized in that: five parts of data that data preparation and preprocessing module generate is subjected to feature selecting according to feature selecting result respectively
The Model Parameter Optimization and training function inputted in the module afterwards carries out model training, and wherein model is by four kinds of different model structures
At respectively logistic regression, random forest, XGBoost and deep learning.
6. a kind of credit customer qualification classification method for passing through machine learning based on WOE conversion according to claim 5,
Be characterized in that: the model deployment module sends client into part customer data into part system, will into part customer data by into
Part data processing module obtains processed customer data, in the model that this partial data input model is loaded to function, finally
Obtain client qualification scoring.
7. a kind of credit customer qualification classification method for passing through machine learning based on WOE conversion according to claim 6,
Be characterized in that: the client qualification scoring that the model deployment module goes out in showing function according to model prediction passes through client qualification
Division module obtains the qualification grade of client, the lending amount of money, manual examination and verification suggestion.
8. a kind of credit customer qualification classification method for passing through machine learning based on WOE conversion according to claim 1,
Be characterized in that: the client qualification division module calculates the accounting of the fine or not qualification client of different segmentations, with reference to the average visitor of history
The lending amount of money, the revenue amount at family calculate the investment return ratio of different segmentations from the loss amount of money, are when referred to according to investment return
Credit product policy assigns different grades of credit standard for the client of different grades, and the client of different segmentations is set
The different lending amount of money and manual examination and verification suggestion.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201910297368.1A CN110322335A (en) | 2019-04-15 | 2019-04-15 | A kind of credit customer qualification classification method passing through machine learning based on WOE conversion |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201910297368.1A CN110322335A (en) | 2019-04-15 | 2019-04-15 | A kind of credit customer qualification classification method passing through machine learning based on WOE conversion |
Publications (1)
Publication Number | Publication Date |
---|---|
CN110322335A true CN110322335A (en) | 2019-10-11 |
Family
ID=68113403
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201910297368.1A Pending CN110322335A (en) | 2019-04-15 | 2019-04-15 | A kind of credit customer qualification classification method passing through machine learning based on WOE conversion |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN110322335A (en) |
Cited By (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN111709828A (en) * | 2020-06-12 | 2020-09-25 | 中国建设银行股份有限公司 | Resource processing method, device, equipment and system |
CN112330440A (en) * | 2020-11-06 | 2021-02-05 | 新华中经信用管理有限公司 | Credit system construction method based on block chain decentralization |
CN113033717A (en) * | 2021-05-26 | 2021-06-25 | 华控清交信息科技(北京)有限公司 | Model generation method and device for model generation |
CN113538131A (en) * | 2021-07-23 | 2021-10-22 | 中信银行股份有限公司 | Method and device for modeling modular scoring card, storage medium and electronic equipment |
CN113610630A (en) * | 2021-08-06 | 2021-11-05 | 东方口岸科技有限公司 | Financial credit modeling method and system based on import and export trade data |
TWI817237B (en) * | 2021-11-04 | 2023-10-01 | 關貿網路股份有限公司 | Method and system for risk prediction and computer-readable medium therefor |
-
2019
- 2019-04-15 CN CN201910297368.1A patent/CN110322335A/en active Pending
Cited By (7)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN111709828A (en) * | 2020-06-12 | 2020-09-25 | 中国建设银行股份有限公司 | Resource processing method, device, equipment and system |
CN112330440A (en) * | 2020-11-06 | 2021-02-05 | 新华中经信用管理有限公司 | Credit system construction method based on block chain decentralization |
CN112330440B (en) * | 2020-11-06 | 2023-10-27 | 新华中经信用管理有限公司 | Credit system construction method based on block chain decentralization |
CN113033717A (en) * | 2021-05-26 | 2021-06-25 | 华控清交信息科技(北京)有限公司 | Model generation method and device for model generation |
CN113538131A (en) * | 2021-07-23 | 2021-10-22 | 中信银行股份有限公司 | Method and device for modeling modular scoring card, storage medium and electronic equipment |
CN113610630A (en) * | 2021-08-06 | 2021-11-05 | 东方口岸科技有限公司 | Financial credit modeling method and system based on import and export trade data |
TWI817237B (en) * | 2021-11-04 | 2023-10-01 | 關貿網路股份有限公司 | Method and system for risk prediction and computer-readable medium therefor |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN110322335A (en) | A kind of credit customer qualification classification method passing through machine learning based on WOE conversion | |
Chang et al. | Application of eXtreme gradient boosting trees in the construction of credit risk assessment models for financial institutions | |
CN108898479B (en) | Credit evaluation model construction method and device | |
CN109409969A (en) | Enterprise tax fraud detection method, electronic equipment and storage medium | |
CN110909984B (en) | Business data processing model training method, business data processing method and device | |
CN112700319A (en) | Enterprise credit line determination method and device based on government affair data | |
CN113537807B (en) | Intelligent wind control method and equipment for enterprises | |
CN108197636A (en) | A kind of paddy detection and sorting technique based on depth multiple views feature | |
CN112700324A (en) | User loan default prediction method based on combination of Catboost and restricted Boltzmann machine | |
CN104346698A (en) | Catering member big data analysis and checking system based on cloud computing and data mining | |
CN104615789A (en) | Data classifying method and device | |
CN110443290A (en) | A kind of product competition relationship quantization generation method and device based on big data | |
CN112037006A (en) | Credit risk identification method and device for small and micro enterprises | |
CN115205026A (en) | Credit evaluation method, device, equipment and computer storage medium | |
CN206897873U (en) | A kind of image procossing and detecting system based on detection product performance | |
CN112465397A (en) | Audit data analysis method and device | |
Daneshmandi et al. | A hybrid data mining model to improve customer response modeling in direct marketing | |
Han | Researches of detection of fraudulent financial statements based on data mining | |
CN115907954A (en) | Account identification method and device, computer equipment and storage medium | |
KR102499182B1 (en) | Loan regular auditing system using artificia intellicence | |
CN114626940A (en) | Data analysis method and device and electronic equipment | |
CN114298472A (en) | Method and system for evaluating images of upstream and downstream enterprises of digital factory | |
Jan et al. | Detection of fraudulent financial statements using decision tree and artificial neural network | |
Meng et al. | The practice study of consumer credit risk based on random forest | |
CN113554278A (en) | Dynamic flexible rule company operation crisis early warning method and system |
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 | ||
WD01 | Invention patent application deemed withdrawn after publication |
Application publication date: 20191011 |
|
WD01 | Invention patent application deemed withdrawn after publication |