CN106960387A - Individual credit risk appraisal procedure and system - Google Patents
Individual credit risk appraisal procedure and system Download PDFInfo
- Publication number
- CN106960387A CN106960387A CN201710296949.4A CN201710296949A CN106960387A CN 106960387 A CN106960387 A CN 106960387A CN 201710296949 A CN201710296949 A CN 201710296949A CN 106960387 A CN106960387 A CN 106960387A
- Authority
- CN
- China
- Prior art keywords
- random forest
- data
- evaluation index
- classification
- result
- 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
Landscapes
- Business, Economics & Management (AREA)
- Accounting & Taxation (AREA)
- Finance (AREA)
- Engineering & Computer Science (AREA)
- Development Economics (AREA)
- Economics (AREA)
- Marketing (AREA)
- Strategic Management (AREA)
- Technology Law (AREA)
- Physics & Mathematics (AREA)
- General Business, Economics & Management (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
Abstract
The invention discloses a kind of personal credit file method and system, wherein method includes:The data of the effective clients of P2P are obtained as raw data set, concentrated using bootstrap methods from initial data randomly select N number of sample data set with putting back to, and build N classification tree, and then N number of sample data set of extraction is put into corresponding classification tree, every classification tree exports a result, random forest is generated according to the data result of all classification trees, finally P2P customer datas are differentiated and classified using random forest, and the individual credit risk of P2P client is estimated with classification results according to differentiating.The present invention can greatly improve the degree of accuracy that user classifies by improved random forest method.
Description
Technical field
The present invention relates to field of computer technology, more particularly to a kind of individual credit risk appraisal procedure and system.
Background technology
Along with China's economic transition to boosting consumption, expanding domestic demand, the active demand for development structure of adjusting economy, and
The lifting of the income of residents and consuming capacity, China's consumptive credit market achieves quick development.But in P2P, (individual is to individual
People) investor subjects huge credit risk in online credit.First, the online credits of majority P2P all do not have in borrowing process
There is mortgage, borrower once breaks one's promise, investor can be caused by huge loss;Second, investor is recognized the information of borrower
Know and come from the online lending platforms of P2P, the asymmetric factor of existence information.So, the assessing credit risks of borrower is that P2P exists
A vital link in line credit, it drastically influence the life cycle of a platform.Therefore, stabilization, efficiently
Assessing credit risks system be particularly important.
At present, general credit evaluation key element is mainly " 5C ", " 5P " and " LAPP " in the world, and main flow business bank is by visitor
The data at family are analyzed user data by some score-systems, the FICO points-scoring systems in such as U.S., quantify the letter of user
With evaluation index, it is weighted finally according to different weights and obtains credit scoring.P2P online credits are due to providing user
Information requirements are not strict, typically only possess the information such as its basic assets information, educational background, age, identity, then pass through third party
Authentication platform to borrower carry out authentification of message, then evaluate borrower credit grade, referred to for investor.
Because the data sample obtained from the online lending platforms of P2P is limited, lack of balance, so, using existing
Online Credit Risk Assessment system is classified to P2P Debit Users, and its precision is relatively low, it is difficult to realize the standard of individual credit risk
Really assess.
The content of the invention
The invention provides a kind of individual credit risk appraisal procedure, comprise the following steps:
S100, the data of the acquisition effective clients of P2P are as raw data set;
S200, using bootstrap methods (bootstrap) from the initial data concentrate randomly select N number of sample with putting back to
Data set, and build N classification tree;
S300, N number of sample data set of extraction is put into corresponding classification tree, every classification tree exports a result,
Random forest is generated according to the data result of all classification trees;
S400, using the random forest P2P customer datas are differentiated and classified, and according to differentiating and classification results
The individual credit risk of the P2P client is estimated.
Wherein, in step S300, N number of sample data set of extraction is put into corresponding classification tree, every classification tree is defeated
Go out a result, random forest is generated according to the data result of all classification trees, comprised the following steps:
S310, to each node randomly choose M evaluation index as feature set to be selected, M is integer;
S320, the selection m (m in the feature set to be selected<M) individual evaluation index calculates its split values Φ (α):
Φ (α)=β1Ginidivide(S)-β2GiniRatio(A)
Wherein, Ginidivide(S) it is:
S1,S2Two subsets being separated into for sample set S;
Gini(S1) measured for CART algorithm partitions:
GiniRatio (A) is the information gain-ratio in improved C4.5 algorithms:
S330, each evaluation index of comparison split values Φ (α), regard the minimum evaluation indexes of split values Φ (α) as section
Dot splitting feature, and delete in feature set to be selected the evaluation index;
Whether the sample that S340, inspection node branch are covered belongs to same class;Same class is such as not belonging to, then basis should
Disruptive features are classified as two subsets, are concentrated in two sons and perform step S310 to S340 successively respectively;Such as belong to same class,
Then generate child node, output category result.
Based on same inventive concept, the present invention also provides a kind of individual credit risk assessment system, including initial data is obtained
Modulus block, data extraction module, random forest generation module and sort module;
The initial data acquisition module, for obtaining the data of the effective clients of P2P as raw data set;
The data extraction module, is taken out with putting back at random for being concentrated using bootstrap methods from the initial data
N number of sample data set is taken, and builds N classification tree;
The random forest generation module, for N number of sample data set of extraction to be put into corresponding classification tree, every
Classification tree exports a result, and random forest is generated according to the data result of all classification trees;
The sort module, for P2P customer datas to be differentiated and classified using the random forest, and according to sentencing
The individual credit risk of the P2P client is not estimated with classification results.
As a kind of embodiment, the random forest generation module includes choosing unit, computing unit, comparing unit
And inspection unit;
The selection unit, for randomly choosing M evaluation index as feature set to be selected to each node, M is integer;
The computing unit, for selecting m (m in the feature set to be selected<M) individual evaluation index calculates its split values Φ
(α):
Φ (α)=β1Ginidivide(S)-β2GiniRatio(A)
Wherein, Ginidivide(S) it is:
S1,S2Two subsets being separated into for sample set S;
Gini(S1) measured for CART algorithm partitions:
GiniRatio (A) is the information gain-ratio in improved C4.5 algorithms:
The comparing unit, the split values Φ (α) for comparing each evaluation index, by commenting for split values Φ (α) minimums
Valency index deletes in feature set to be selected the evaluation index as node split feature;
The inspection unit, for checking whether the sample that node branch is covered belongs to same class;Such as it is not belonging to same
One class, then be classified as two subsets according to the disruptive features, is concentrated in two sons and perform successively respectively selection unit, calculates single
Member, the action of comparing unit;Such as belong to same class, then generate child node, output category result.
The present invention is compared to the beneficial effect of prior art:
Individual credit risk appraisal procedure and system that the present invention is provided, by obtain the data of the effective clients of P2P by its
As raw data set, concentrated using bootstrap methods from initial data randomly select N number of sample data set with putting back to, and
N classification tree is built, and then N number of sample data set of extraction is put into corresponding classification tree, every classification tree exports a knot
Really, random forest is generated according to the data result of all classification trees, finally P2P customer datas sentenced using random forest
Other and classification, and the individual credit risk of P2P client is estimated with classification results according to differentiating.The present invention is by improved
Random forest method can greatly improve the degree of accuracy of user's classification.
Brief description of the drawings
The schematic flow sheet for the individual credit risk appraisal procedure that Fig. 1 provides for one embodiment of the invention;
Fig. 2 is the principle schematic of the individual credit risk appraisal procedure shown in Fig. 1;
Fig. 3 illustrates for the flow of an embodiment of the step S300 in the individual credit risk appraisal procedure shown in Fig. 1
Figure;
The principle schematic for the individual credit risk assessment system that Fig. 4 provides for another embodiment of the present invention.
Embodiment
Below in conjunction with accompanying drawing, the technical characteristic above-mentioned and other to the present invention and advantage are clearly and completely described,
Obviously, described embodiment is only the section Example of the present invention, rather than whole embodiments.
Random forests algorithm has high accuracy rate and the tolerance good to lack of balance sample and noise, so
Random forest method is introduced into the online Credit Risk Assessments of P2P by the present invention.Random forest Method Modeling thought is to utilize
Bootstrap methods are randomly selected from original sample set obtains multiple subsample collection, and decision-making is carried out to each subsample collection
Tree modeling, then according to ballot method predicting the outcome to be voted and determine predicting the outcome for random forest to many decision trees.
Refer to Fig. 1 and Fig. 2, the individual credit risk appraisal procedure that one embodiment of the invention is provided, the invention provides
A kind of individual credit risk appraisal procedure, comprises the following steps:
S100, the data of the acquisition effective clients of P2P are as raw data set;
S200, concentrated using bootstrap from initial data randomly select N number of sample data set with putting back to, and build N point
Class tree;
S300, N number of sample data set of extraction is put into corresponding classification tree, every classification tree exports a result,
Random forest is generated according to the data result of all classification trees;
S400, using random forest P2P customer datas are differentiated and classified, and according to differentiating and classification results pair
The individual credit risk of P2P client is estimated.
As a kind of embodiment, in step S300, N number of sample data set of extraction is put into corresponding classification tree
In, every classification tree exports a result, and random forest, including following step are generated according to the data result of all classification trees
Suddenly:
S310, to each node randomly choose M evaluation index as feature set to be selected, M is integer;
S320, the selection m (m in feature set to be selected<M) individual evaluation index calculates its split values Φ (α):
Φ (α)=β1Ginidivide(S)-β2GiniRatio(A)
Wherein, Ginidivide(S) it is:
S1,S2Two subsets being separated into for sample set S;
Gini(S1) measured for CART algorithm partitions:
GiniRatio (A) is the information gain-ratio in improved C4.5 algorithms:
S330, each evaluation index of comparison split values Φ (α), regard the minimum evaluation indexes of split values Φ (α) as section
Dot splitting feature, and delete in feature set to be selected the evaluation index;
Whether the sample that S340, inspection node branch are covered belongs to same class;Same class is such as not belonging to, then basis should
Disruptive features are classified as two subsets of A, B, are concentrated in two sons of A, B and perform step S310 to S340 successively respectively;Such as belong to
Same class, then generate child node, output category result.
Will extract in the individual credit risk appraisal procedure provided referring to Fig. 3, another embodiment of the present invention, step S300
N number of sample data set be put into corresponding classification tree, every classification tree exports a result, can pass through following steps real
It is existing:
S301, m evaluation index of random selection obtain feature set to be selected;
S302, select n evaluation index in feature set to be selected and calculate its split values;
S303, it regard the minimum evaluation index of split values as node split feature;
S304, the evaluation index is deleted in feature set to be selected;
Whether the sample that S305, the branch of decision node are covered belongs to same class, if it is not, then return to step S302;
S306, if so, then generate child node, output category result.
Based on same inventive concept, the present invention also provides a kind of individual credit risk assessment system, the system and above-mentioned side
The principle of method is identical, and the implementation of system can refer to above method realization, repeat part no longer redundant later.
Referring to Fig. 4, the individual credit risk assessment system that the present invention is provided includes initial data acquisition module 100, data
Abstraction module 200, random forest generation module 300 and sort module 400.Wherein, initial data acquisition module 100 is used to obtain
The data of the effective clients of P2P are taken as raw data set;Data extraction module 200 is used to use bootstrap methods from original
Beginning data are concentrated with randomly selecting N number of sample data set with putting back to, and build N classification tree;Random forest generation module 300 is used for
N number of sample data set of extraction is put into corresponding classification tree, every classification tree exports a result, according to all classification
The data result generation random forest of tree;Sort module 400 is used to P2P customer datas are differentiated and divided using random forest
Class, and the individual credit risk of P2P client is estimated with classification results according to differentiating.
As a kind of embodiment, random forest generation module 300 includes choosing unit, computing unit, comparing unit
And inspection unit.Wherein:
Choosing unit is used to randomly choose M evaluation index as feature set to be selected to each node, and M is integer.
Computing unit is used in feature set to be selected select m (m<M) individual evaluation index calculates its split values Φ (α):
Φ (α)=β1Ginidivide(S)-β2GiniRatio(A)
Wherein, Ginidivide(S) it is:
S1,S2Two subsets being separated into for sample set S;
Gini(S1) measured for CART algorithm partitions:
GiniRatio (A) is the information gain-ratio in improved C4.5 algorithms:
Comparing unit is used for the split values Φ (α) for comparing each evaluation index, by the evaluation index that split values Φ (α) is minimum
As node split feature, and delete in feature set to be selected the evaluation index.
Inspection unit is used to check whether the sample that node branch is covered belongs to same class;Same class is such as not belonging to,
Two subsets then are classified as according to the disruptive features, concentrates to perform successively respectively in two sons and chooses unit, computing unit, ratio
Compared with the action of unit;Such as belong to same class, then generate child node, output category result.
Individual credit risk appraisal procedure and system that the present invention is provided, by obtain the data of the effective clients of P2P by its
As raw data set, concentrated using bootstrap methods from the initial data randomly select N number of sample data with putting back to
Collection, and N classification tree is built, and then N number of sample data set of extraction is put into corresponding classification tree, every classification tree output one
Individual result, generates random forest, finally using the random forest to P2P client's number according to the data result of all classification trees
According to being differentiated and classified, and according to differentiating and classification results are estimated to the individual credit risk of the P2P client.This hair
It is bright that the degree of accuracy that user classifies can be greatly improved by improved random forest method.
Particular embodiments described above, has been carried out further to the purpose of the present invention, technical scheme and beneficial effect
Describe in detail, it will be appreciated that the foregoing is only the specific embodiment of the present invention, the protection being not intended to limit the present invention
Scope.Particularly point out, to those skilled in the art, within the spirit and principles of the invention, that is done any repaiies
Change, equivalent, improvement etc., should be included within the scope of the present invention.
Claims (4)
1. a kind of individual credit risk appraisal procedure, it is characterised in that comprise the following steps:
S100, the data of the acquisition effective clients of P2P are as raw data set;
S200, concentrated using bootstrap methods from the initial data randomly select N number of sample data set with putting back to, and build N
Classification tree;
S300, N number of sample data set of extraction is put into corresponding classification tree, every classification tree exports a result, according to
The data result generation random forest of all classification trees;
S400, using the random forest P2P customer datas are differentiated and classified, and according to differentiating and classification results are to institute
The individual credit risk for stating P2P client is estimated.
2. individual credit risk appraisal procedure according to claim 1, it is characterised in that in step S300, by the N of extraction
Individual sample data set is put into corresponding classification tree, and every classification tree exports a result, according to the data of all classification trees
As a result random forest is generated, is comprised the following steps:
S310, to each node randomly choose M evaluation index as feature set to be selected, M is integer;
S320, the selection m (m in the feature set to be selected<M) individual evaluation index calculates its split values Φ (α):
Φ (α)=β1Ginidivide(S)-β2GiniRatio(A)
Wherein, Ginidivide(S) it is:
S1,S2Two subsets being separated into for sample set S;
Gini(S1) measured for CART algorithm partitions:
GiniRatio (A) is the information gain-ratio in improved C4.5 algorithms:
S330, each evaluation index of comparison split values Φ (α), regard evaluation index minimum split values Φ (α) as node point
Feature is split, and deletes in feature set to be selected the evaluation index;
Whether the sample that S340, inspection node branch are covered belongs to same class;Same class is such as not belonging to, then according to the division
Feature is classified as two subsets, is concentrated in two sons and performs step S310 to S340 successively respectively;Such as belong to same class, then give birth to
Into child node, output category result.
3. a kind of individual credit risk assessment system, it is characterised in that including initial data acquisition module, data extraction module,
Random forest generation module and sort module;
The initial data acquisition module, for obtaining the data of the effective clients of P2P as raw data set;
The data extraction module, it is N number of for being randomly selected with putting back to from initial data concentration using bootstrap methods
Sample data set, and build N classification tree;
The random forest generation module, for N number of sample data set of extraction to be put into corresponding classification tree, every classification
Tree one result of output, random forest is generated according to the data result of all classification trees;
The sort module, for P2P customer datas to be differentiated and classified using the random forest, and according to differentiate and
Classification results are estimated to the individual credit risk of the P2P client.
4. individual credit risk assessment system according to claim 3, it is characterised in that the random forest generation module
Including choosing unit, computing unit, comparing unit and inspection unit;
The selection unit, for randomly choosing M evaluation index as feature set to be selected to each node, M is integer;
The computing unit, for selecting m (m in the feature set to be selected<M) individual evaluation index calculates its split values Φ (α):
Φ (α)=β1Ginidivide(S)-β2GiniRatio(A)
Wherein, Ginidivide(S) it is:
S1,S2Two subsets being separated into for sample set S;
Gini(S1) measured for CART algorithm partitions:
GiniRatio (A) is the information gain-ratio in improved C4.5 algorithms:
The comparing unit, the split values Φ (α) for comparing each evaluation index refers to the minimum evaluations of split values Φ (α)
It is denoted as node split feature, and deletes in feature set to be selected the evaluation index;
The inspection unit, for checking whether the sample that node branch is covered belongs to same class;Same class is such as not belonging to,
Two subsets then are classified as according to the disruptive features, concentrates to perform successively respectively in two sons and chooses unit, computing unit, ratio
Compared with the action of unit;Such as belong to same class, then generate child node, output category result.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201710296949.4A CN106960387A (en) | 2017-04-28 | 2017-04-28 | Individual credit risk appraisal procedure and system |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201710296949.4A CN106960387A (en) | 2017-04-28 | 2017-04-28 | Individual credit risk appraisal procedure and system |
Publications (1)
Publication Number | Publication Date |
---|---|
CN106960387A true CN106960387A (en) | 2017-07-18 |
Family
ID=59484653
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201710296949.4A Pending CN106960387A (en) | 2017-04-28 | 2017-04-28 | Individual credit risk appraisal procedure and system |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN106960387A (en) |
Cited By (9)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN108960561A (en) * | 2018-05-04 | 2018-12-07 | 阿里巴巴集团控股有限公司 | A kind of air control model treatment method, device and equipment based on unbalanced data |
CN109325844A (en) * | 2018-06-25 | 2019-02-12 | 南京工业大学 | Net under multidimensional data borrows borrower's credit assessment method |
CN109767317A (en) * | 2018-12-15 | 2019-05-17 | 深圳壹账通智能科技有限公司 | Loan checking method, device, equipment and medium based on membership grade evaluation |
CN109903140A (en) * | 2019-03-07 | 2019-06-18 | 阿里巴巴集团控股有限公司 | A kind of credit services recommended method, device and equipment |
WO2019120023A1 (en) * | 2017-12-22 | 2019-06-27 | Oppo广东移动通信有限公司 | Gender prediction method and apparatus, storage medium and electronic device |
CN110334737A (en) * | 2019-06-04 | 2019-10-15 | 阿里巴巴集团控股有限公司 | A kind of method and system of the customer risk index screening based on random forest |
CN110443692A (en) * | 2019-07-04 | 2019-11-12 | 平安科技(深圳)有限公司 | Enterprise's credit authorization method, apparatus, equipment and computer readable storage medium |
CN110827131A (en) * | 2018-07-23 | 2020-02-21 | 中国软件与技术服务股份有限公司 | Tax payer credit evaluation method based on distributed automatic feature combination |
CN110826618A (en) * | 2019-11-01 | 2020-02-21 | 南京信息工程大学 | Personal credit risk assessment method based on random forest |
-
2017
- 2017-04-28 CN CN201710296949.4A patent/CN106960387A/en active Pending
Cited By (13)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
WO2019120023A1 (en) * | 2017-12-22 | 2019-06-27 | Oppo广东移动通信有限公司 | Gender prediction method and apparatus, storage medium and electronic device |
CN109961077A (en) * | 2017-12-22 | 2019-07-02 | 广东欧珀移动通信有限公司 | Gender prediction's method, apparatus, storage medium and electronic equipment |
CN108960561A (en) * | 2018-05-04 | 2018-12-07 | 阿里巴巴集团控股有限公司 | A kind of air control model treatment method, device and equipment based on unbalanced data |
CN109325844A (en) * | 2018-06-25 | 2019-02-12 | 南京工业大学 | Net under multidimensional data borrows borrower's credit assessment method |
CN110827131A (en) * | 2018-07-23 | 2020-02-21 | 中国软件与技术服务股份有限公司 | Tax payer credit evaluation method based on distributed automatic feature combination |
CN110827131B (en) * | 2018-07-23 | 2022-06-28 | 中国软件与技术服务股份有限公司 | Tax payer credit evaluation method based on distributed automatic feature combination |
CN109767317A (en) * | 2018-12-15 | 2019-05-17 | 深圳壹账通智能科技有限公司 | Loan checking method, device, equipment and medium based on membership grade evaluation |
CN109903140A (en) * | 2019-03-07 | 2019-06-18 | 阿里巴巴集团控股有限公司 | A kind of credit services recommended method, device and equipment |
CN110334737A (en) * | 2019-06-04 | 2019-10-15 | 阿里巴巴集团控股有限公司 | A kind of method and system of the customer risk index screening based on random forest |
CN110334737B (en) * | 2019-06-04 | 2023-04-07 | 创新先进技术有限公司 | Customer risk index screening method and system based on random forest |
CN110443692A (en) * | 2019-07-04 | 2019-11-12 | 平安科技(深圳)有限公司 | Enterprise's credit authorization method, apparatus, equipment and computer readable storage medium |
CN110443692B (en) * | 2019-07-04 | 2024-05-10 | 平安科技(深圳)有限公司 | Enterprise credit auditing method, device, equipment and computer readable storage medium |
CN110826618A (en) * | 2019-11-01 | 2020-02-21 | 南京信息工程大学 | Personal credit risk assessment method based on random forest |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN106960387A (en) | Individual credit risk appraisal procedure and system | |
CN110334737B (en) | Customer risk index screening method and system based on random forest | |
Abdou et al. | Predicting creditworthiness in retail banking with limited scoring data | |
CN108665159A (en) | A kind of methods of risk assessment, device, terminal device and storage medium | |
CN111311416B (en) | Block chain money laundering node detection method based on multichannel graph and graph neural network | |
CN103678659A (en) | E-commerce website cheat user identification method and system based on random forest algorithm | |
CN103886108B (en) | The feature selecting and weighing computation method of a kind of unbalanced text set | |
Mathew et al. | Perceptions and intentions of customers towards mobile banking adoption | |
Habiba et al. | A comparative study on fake job post prediction using different data mining techniques | |
CN110084609A (en) | A kind of transaction swindling behavior depth detection method based on representative learning | |
CN105869058A (en) | Method for user portrait extraction based on multilayer latent variable model | |
Verma et al. | Credit Card Fraud Detection Using Selective Class Sampling and Random Forest Classifier | |
Liu et al. | An innovative model fusion algorithm to improve the recall rate of peer-to-peer lending default customers | |
Chen et al. | Mixed credit scoring model of logistic regression and evidence weight in the background of big data | |
CN114298834A (en) | Personal credit evaluation method and system based on self-organizing mapping network | |
Yangyudongnanxin | Financial credit risk control strategy based on weighted random forest algorithm | |
Ozgulbas et al. | Financial profiling of public hospitals: an application by data mining | |
CN109801162A (en) | A kind of credit rating method that social media data are merged with multi-standard cross-certification | |
Wijesoma et al. | Selecting optimal personalized features for on-line signature verification using GA | |
Zhang | Research on credit risk forecast model based on data mining technology | |
CN113706258A (en) | Product recommendation method, device, equipment and storage medium based on combined model | |
Faraj et al. | Comparison of different ensemble methods in credit card default prediction | |
Reddy et al. | Peer to peer lending, default prediction-evidence from lending club | |
Chi et al. | A hybrid model for credit risk assessment: empirical validation by real-world credit data | |
Raj et al. | Automated Cyberstalking Classification using Social Media |
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 | ||
RJ01 | Rejection of invention patent application after publication |
Application publication date: 20170718 |