CN110837963A - Risk control platform construction method based on data, model and strategy - Google Patents
Risk control platform construction method based on data, model and strategy Download PDFInfo
- Publication number
- CN110837963A CN110837963A CN201911074312.6A CN201911074312A CN110837963A CN 110837963 A CN110837963 A CN 110837963A CN 201911074312 A CN201911074312 A CN 201911074312A CN 110837963 A CN110837963 A CN 110837963A
- Authority
- CN
- China
- Prior art keywords
- data
- different
- strategy
- configuration
- management module
- 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.)
- Withdrawn
Links
Images
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
- G06Q10/00—Administration; Management
- G06Q10/06—Resources, workflows, human or project management; Enterprise or organisation planning; Enterprise or organisation modelling
- G06Q10/063—Operations research, analysis or management
- G06Q10/0635—Risk analysis of enterprise or organisation activities
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/20—Information retrieval; Database structures therefor; File system structures therefor of structured data, e.g. relational data
- G06F16/25—Integrating or interfacing systems involving database management systems
- G06F16/252—Integrating or interfacing systems involving database management systems between a Database Management System and a front-end application
Abstract
The invention discloses a risk control platform construction method based on data, models and strategies, and relates to the technical field of data control; and (3) setting up a risk control platform: the data management module is established and is responsible for acquiring and processing data through a data source, and meanwhile, a data warehouse is utilized to establish different data sets according to different data types for data storage and maintenance; establishing a model calculation module, selecting different models for analysis and prediction according to the data of the user and different services, and obtaining the result data of the analysis and prediction; establishing a strategy management module, performing strategy configuration on the existing original data and the result data subjected to analysis and prediction according to the service requirement, utilizing a rule engine to check the result data of the strategy configuration one by one, feeding back the result data of the strategy configuration to form different strategy execution results, and distinguishing the user credit; and establishing a financial product supervision module, configuring corresponding financial products according to the user credit degree, and supervising the user.
Description
Technical Field
The invention discloses a risk control platform construction method based on data, models and strategies, and relates to the technical field of data control.
Background
Risk control refers to the risk manager taking various measures and methods to eliminate or reduce the various possibilities of occurrence of a risk event, or the risk controller reducing the losses incurred when a risk event occurs. As a manager, various measures are taken to reduce the possibility of occurrence of the risk event or control the possible loss within a certain range so as to avoid the loss which is hard to bear when the risk event occurs.
The existing bank risk control mode cannot effectively meet the increasingly complex credit investigation market requirement due to the adverse factors of time lag, incapability of realizing full online process, low user portrait accuracy, small control coverage and the like.
The invention provides a risk control platform construction method based on data, models and strategies, which comprises the following main construction flow steps: by butt joint of public data, self data, three-party access data, risk data and the like, various data sources are gathered on a data management platform, and data transmission, storage, management, maintenance and visualization are carried out. Under the condition that real-time data and historical storage data are complete, technologies such as big data and machine learning are adopted, and intelligent, real-time and efficient data processing and modeling capabilities are given to the risk control platform. The model construction process is completed in a process, a one-stop mode and an online mode from data extraction, data processing, data analysis, feature training, model evaluation and model training to model deployment and iterative optimization. The risk control policy is a configurable business process based on raw data or model outcome data. The risk control strategy comprises an anti-fraud strategy, a scoring strategy, a credit granting strategy and the like, and each strategy has corresponding business logic. And acquiring information data of a corresponding lender by calling the configured data source, and carrying out rule verification or logic judgment on the data to generate a final result of the current strategy. The combination of different strategies forms different credit products which can be output externally, and the credit decision service is uniformly provided externally under the condition that the platform access authority is met. The risk control platform also comprises a post-loan monitoring platform which periodically scans overdue records after the customer loan and establishes risk control standards of different levels for different customers so as to realize all-round prevention and control before, during and after the loan.
Disclosure of Invention
Aiming at the problems in the prior art, the invention provides a risk control platform construction method based on data, models and strategies, provides an output flow of a risk control platform, provides a method for solving the problem that credit products output risk control results on a pure line through a financial risk control platform, and can be popularized widely based on the credit products.
The specific scheme provided by the invention is as follows:
a risk control platform construction method based on data, models and strategies,
and (3) setting up a risk control platform:
the data management module is established and is responsible for acquiring and processing data through a data source, and meanwhile, a data warehouse is utilized to establish different data sets according to different data types for data storage and maintenance;
establishing a model calculation module, selecting different models for analyzing and predicting according to the data of the user and different services to obtain the result data of the analysis and prediction,
establishing a policy management module, performing policy configuration on the existing original data and the analyzed and predicted result data according to the service requirement, utilizing a rule engine to check the result data of the policy configuration one by one, feeding back the result data of the policy configuration to form different policy execution results, distinguishing the user credit degrees,
and establishing a financial product supervision module, configuring corresponding financial products according to the user credit degree, and supervising the user.
In the data management module established in the method, when the data source is diversified, the data source is acquired in different modes according to the data source.
The method adopts the mode of collecting data from a database, collecting logs or burying points in pages to collect data of different data sources.
In the method, a model in a model calculation module is trained and tested through a data set collected in a data management module to obtain a mature model.
The model calculation module established in the method comprises an enterprise anti-fraud model based on the blacklist library and the relation map, an individual industrial and commercial customer rating card model based on legal basic information data and enterprise operation data, and a quota measurement model based on personal social security and accumulation fund data.
In the method, a strategy management module carries out related configuration on the existing original data and the result data which is analyzed and predicted according to an anti-fraud strategy, a scoring strategy and a credit granting strategy.
The related configuration in the method comprises the configuration of a data source, the configuration of an admission rule, the configuration of a truth-verifying condition and the configuration of an anti-fraud rule.
In the method, a rule engine checks result data configured by the strategies one by using a Json Array format, and the basic syntax form is as follows: and more complex judgment conditions can be realized by expanding expressions.
A risk control platform based on data, model and strategy comprises a data management module, a model calculation module, a strategy management module and a financial product supervision module,
the data management module is responsible for acquiring and processing data through a data source, and meanwhile, a data warehouse is used for establishing different data sets according to different data types to store and maintain the data;
the model calculation module selects different models for analysis and prediction according to the data of the user and different services to obtain the result data of the analysis and prediction,
the strategy management module carries out strategy configuration on the existing original data and the result data which is analyzed and predicted according to the service requirement, utilizes the rule engine to carry out one-by-one inspection on the result data of the strategy configuration, feeds back the result data of the strategy configuration to form different strategy execution results, distinguishes the user credit degree,
and the financial product supervision module is used for configuring corresponding financial products according to the user credit and supervising the user.
The invention has the advantages that:
the invention provides a risk control platform construction method based on data, models and strategies, which is used for establishing a risk control platform and has the following beneficial results:
through big data drive, the defects of limited data source, simple principle and large risk of the conventional financial product process are overcome by the advantages of various data, compatibility with each platform and full online result output;
a large number of risk control models and product strategies are preset, various service scenes can be quickly matched, different user requirements are met, and compared with the existing risk control technology, the risk control efficiency is improved, and the credit reject ratio is reduced;
the method is more suitable for continuously changing service scenes, continuously carries out iteration and dynamic adjustment on user data, and can continuously improve the evaluation result of the risk control product by continuously adjusting and optimizing product strategy indexes and verification standards according to samples;
the intelligent risk control platform based on artificial intelligence and various platform framework technologies can be introduced in the later stage, version updating can be rapidly realized by the aid of the platform due to the characteristic of high expandability, financial product users with high risks can be identified more accurately after updating, and the application prospect is wider.
Drawings
FIG. 1 is a schematic diagram of the data flow of the method of the present invention;
FIG. 2 is a schematic diagram of a platform architecture according to the present invention;
FIG. 3 is a schematic diagram of a specific application of the method of the present invention.
Detailed Description
The invention provides a risk control platform construction method based on data, models and strategies,
and (3) setting up a risk control platform:
the data management module is established and is responsible for acquiring and processing data through a data source, and meanwhile, a data warehouse is utilized to establish different data sets according to different data types for data storage and maintenance;
establishing a model calculation module, selecting different models for analyzing and predicting according to the data of the user and different services to obtain the result data of the analysis and prediction,
establishing a policy management module, performing policy configuration on the existing original data and the analyzed and predicted result data according to the service requirement, utilizing a rule engine to check the result data of the policy configuration one by one, feeding back the result data of the policy configuration to form different policy execution results, distinguishing the user credit degrees,
and establishing a financial product supervision module, configuring corresponding financial products according to the user credit degree, and supervising the user.
Meanwhile, a risk control platform based on data, models and strategies corresponding to the method is provided, which comprises a data management module, a model calculation module, a strategy management module and a financial product supervision module,
the data management module is responsible for acquiring and processing data through a data source, and meanwhile, a data warehouse is used for establishing different data sets according to different data types to store and maintain the data;
the model calculation module selects different models for analysis and prediction according to the data of the user and different services to obtain the result data of the analysis and prediction,
the strategy management module carries out strategy configuration on the existing original data and the result data which is analyzed and predicted according to the service requirement, utilizes the rule engine to carry out one-by-one inspection on the result data of the strategy configuration, feeds back the result data of the strategy configuration to form different strategy execution results, distinguishes the user credit degree,
and the financial product supervision module is used for configuring corresponding financial products according to the user credit and supervising the user.
The present invention is further described below in conjunction with the following figures and specific examples so that those skilled in the art may better understand the present invention and practice it, but the examples are not intended to limit the present invention.
The method is used for constructing a risk control platform based on data, models and strategies, and comprises the following specific processes:
the method comprises the steps of establishing a data management module, wherein the data management module can be divided into internal data and external data from the source, and can be divided into structured and unstructured data, original data and calculation data, discrete data and contact data, black and red list data, portrait data, knowledge map data and the like according to the characteristics of the data. Because the knowledge graph data is data based on a graph data structure, neo4j or integrated Spark graph is mainly used for storage and processing, and with the increase of traffic, Hadoop + Spark can be used for data management and maintenance;
establishing a model calculation module, wherein the model calculation module comprises common algorithm models in risk control business, such as an enterprise anti-fraud model based on a blacklist library and a relational graph, an individual industrial and commercial customer rating card model based on corporate basic information data and enterprise operation data, a quota measurement model based on personal social security and accumulation fund data, and the like, each model can be embedded with algorithm models, such as decision trees, logistic regression, random forests, SVM and the like, according to different model characteristics and business capabilities, and is trained and tested through a data set collected in a data management module to obtain a mature model, such as a rating card model, which can extract data resources stored in a data warehouse, clean and pre-process the data, pay attention to the characteristic division, abnormal values and continuous variable segmentation of the data, select characteristic variables, and pay attention to the characteristic division, abnormal values and continuous variable segmentation of the data in Tensorflow, and select the Tensorlow flow, Dividing samples into a training set and a test set in a GDBT (generalized binary bit-rate) frame environment, performing parameter tuning, entering a model prediction stage, and expressing model discrimination capacity by using a GINI (generalized exponential decay) coefficient, generalization capacity by using an ROC (rock characteristic) curve and the like;
analyzing and predicting the user data and different services by utilizing a mature model to obtain the result data of the analysis and prediction,
establishing a policy management module, performing policy configuration on the existing original data and the analyzed and predicted result data according to the service requirement, wherein the policy comprises an anti-fraud policy, a scoring policy, a credit granting policy and the like, the configuration comprises the configuration of a data source, the configuration of an admission rule, the configuration of a verification condition, the configuration of an anti-fraud rule and the like,
and the result data of the strategy configuration is checked one by utilizing a rule engine, different strategy execution results are formed by feeding back the result data of the strategy configuration, the credit degrees of users are distinguished,
the rule grammar of the rule engine is realized by using the format of Json Array, and the basic grammar form is as follows: [ "operator", "parameter 1", "parameter 2" ], the more complicated judgement condition can be realized by expanding the expression:
[ "operator",
[ "operator 1", "parameter 2",
]
execution with the above expression yields the check result for the current policy, True indicates pass, False indicates fail, and there are other numeric results, and the operator of the supportable operation is shown in table 1,
TABLE 1
After all the verification is completed, different strategy execution results are generated. Of course, different strategies have different characteristics, the scoring strategy can calculate the weight accumulation of a plurality of scoring cards according to different models, the rating strategy can configure credit rating coefficients according to the credit rating of the scoring result, and the like,
the method comprises the steps of establishing a financial product supervision module, managing users and credit products such as financial institutions and the like, configuring different rule items based on the existing strategy set, model result data and data source data, generating different strategies, and finally generating the credit products which can be issued externally, and giving feedback results of the credit application products selected by the current credit applicant, including whether the credit application is refused, credit rating, scoring results, limit suggestions and the like. An overall wind control report may also be given based on the user's characteristic portraits,
meanwhile, the user is supervised, overdue conditions, fund transfer, operation conditions and the like of a loan application subject are continuously explored, a staged wind control result is generated, when various indexes generate large fluctuation to send overdue alarms, the credit line is degraded by adopting a certain means, and when the loss can not be recovered, a post-loan collection mechanism is started, the information related to the loan application person can enter a black list library for data analysis of a later model, and when the loan application person entering the black list library applies the loan again, the credit level of the loan application person can be greatly reduced.
The platform of the invention is used for risk control, and the specific process is as follows:
the method comprises the steps of utilizing a data management module to collect data, wherein the data can be divided into internal data and external data from the source, according to the characteristics of the data, the data can be divided into structured data and unstructured data, original data and calculation data, discrete data and contact data, black and red list data, portrait data, knowledge map data and the like, according to the difference of data sources, the data can be collected from a database, collected in a log mode and buried in a page, and the data are stored in a data warehouse, for example, the external data can be called in a data product interface mode and then directly stored in an Oracle database, the data in the log mode can be collected through flash, output to Kafka and aggregated to an ASELtic for storage, and the black and red list data are stored by using a memory database Redis. Because the knowledge graph data is data based on a graph data structure, neo4j or integrated Spark graph is mainly used for storage and processing, and with the increase of traffic, Hadoop + Spark can be used for data management and maintenance;
the model calculation module comprises common algorithm models in risk control business, such as enterprise anti-fraud models based on black name list library and relational map, individual industry and business evaluation card models based on legal basic information data and enterprise operation data, and limit measurement models based on personal social security and accumulation fund data, according to different model characteristics and business capabilities, each model can be embedded with algorithm models, such as decision trees, logistic regression, random forests, SVM and the like, and is trained and tested through a data set collected in the data management module to obtain mature models, such as evaluation card models, which can extract data resources stored in a data warehouse, clean and pre-process characteristics of data, pay attention to characteristic division, abnormal values and continuous variable segmentation of the data, select characteristic variables, divide samples into training and testing sets in a frame environment such as Tensorflow and GDBT and the like, after parameter tuning, entering a model prediction stage, using a GINI coefficient to represent the discrimination capability of the model, using an ROC curve to represent the generalization capability of the model and the like;
analyzing and predicting the user data and different services by utilizing a mature model to obtain the result data of the analysis and prediction,
the strategy management module carries out strategy configuration on the existing original data and the result data after analysis and prediction according to the service requirement, the strategies comprise anti-fraud strategies, scoring strategies, credit strategies and the like, the configuration comprises the configuration of data sources, the configuration of admission rules, the configuration of verification conditions, the configuration of anti-fraud rules and the like,
and the result data of the strategy configuration is checked one by utilizing a rule engine, different strategy execution results are formed by feeding back the result data of the strategy configuration, the credit degrees of users are distinguished,
the rule grammar of the rule engine is realized by using the format of Json Array, and the basic grammar form is as follows: [ "operator", "parameter 1", "parameter 2" ], the more complicated judgement condition can be realized by expanding the expression:
[ "operator",
[ "operator 1", "parameter 2",
]
execution with the above expression yields the check result for the current policy, True indicates pass, False indicates fail, and there are other numeric results, and the operator of the supportable operation is shown in table 1,
TABLE 1
After all the verification is completed, different strategy execution results are generated. Of course, different strategies have different characteristics, the scoring strategy can calculate the weight accumulation of a plurality of scoring cards according to different models, the rating strategy can configure credit rating coefficients according to the credit rating of the scoring result, and the like,
the financial product supervision module can manage users and credit products such as financial institutions and the like, different rule items are configured firstly based on the existing strategy set, model result data and data source data, after different strategies are generated, the credit products which can be issued to the outside are generated finally, and feedback results of the loan application products selected by the current loan application person can be given, including whether the loan application is refused, credit levels, scoring results, limit suggestions and the like. An overall wind control report may also be given based on the user's characteristic portraits,
meanwhile, the user is supervised, overdue conditions, fund transfer, operation conditions and the like of a loan application subject are continuously explored, a staged wind control result is generated, when various indexes generate large fluctuation to send overdue alarms, the credit line is degraded by adopting a certain means, and when the loss can not be recovered, a post-loan collection mechanism is started, the information related to the loan application person can enter a black list library for data analysis of a later model, and when the loan application person entering the black list library applies the loan again, the credit level of the loan application person can be greatly reduced.
The method or the platform can be used for transmitting, storing, managing, analyzing and visualizing the data of each data source. And (3) carrying out data cleaning and processing on each data source data through a big data stream type processing technology to construct a data warehouse. And building a model and calculating based on big data and machine learning domain knowledge. Both the model result data and the data source data can be used as rule configuration data sources of the strategy set. Different policy combinations will constitute different credit products, unify the external credit provision decision service. After customers of credit product service grow, user groups of different levels are generated according to the post-credit monitoring platform, different characteristic divisions of various user groups are stored in the data warehouse and applied to the construction of the model system, and the purposes of full-flow and all-around prevention and control are achieved.
The above-mentioned embodiments are merely preferred embodiments for fully illustrating the present invention, and the scope of the present invention is not limited thereto. The equivalent substitution or change made by the technical personnel in the technical field on the basis of the invention is all within the protection scope of the invention. The protection scope of the invention is subject to the claims.
Claims (9)
1. A risk control platform construction method based on data, models and strategies is characterized by comprising the following steps of:
the data management module is established and is responsible for acquiring and processing data through a data source, and meanwhile, a data warehouse is utilized to establish different data sets according to different data types for data storage and maintenance;
establishing a model calculation module, selecting different models for analyzing and predicting according to the data of the user and different services to obtain the result data of the analysis and prediction,
establishing a policy management module, performing policy configuration on the existing original data and the analyzed and predicted result data according to the service requirement, utilizing a rule engine to check the result data of the policy configuration one by one, feeding back the result data of the policy configuration to form different policy execution results, distinguishing the user credit degrees,
and establishing a financial product supervision module, configuring corresponding financial products according to the user credit degree, and supervising the user.
2. The method as claimed in claim 1, wherein the data management module is configured to collect the data sources in different manners according to the data sources when the data sources are diversified.
3. The method of claim 2, wherein the data from the different data sources is collected by collection from a database, log collection, or page spotting.
4. A method according to any of claims 1-3, characterized in that the model in the model calculation module is trained and tested by means of the data sets collected in the data management module to obtain a mature model.
5. The method as claimed in claim 4, wherein the established model calculation module comprises an enterprise anti-fraud model based on the blacklist library and the relationship map, an individual industrial and commercial customer rating card model based on the corporate basic information data and the enterprise operation data, and a credit measurement model based on the personal social security and the accumulation fund data.
6. The method as claimed in claim 1 or 5, wherein the policy management module performs related configuration on the existing raw data and the analyzed and predicted result data according to an anti-fraud policy, a scoring policy and a credit granting policy.
7. The method according to claim 6, characterized in that the relevant configuration comprises configuration of data sources, configuration of admission rules, configuration of authentication conditions, configuration of anti-fraud rules.
8. A method as claimed in claim 1 or 7, wherein the rules engine performs a one-by-one check on the policy configured result data using the format of the Json Array, the basic syntax being in the form of: and more complex judgment conditions can be realized by expanding expressions.
9. A risk control platform based on data, models and strategies is characterized by comprising a data management module, a model calculation module, a strategy management module and a financial product supervision module,
the data management module is responsible for acquiring and processing data through a data source, and meanwhile, a data warehouse is used for establishing different data sets according to different data types to store and maintain the data;
the model calculation module selects different models for analysis and prediction according to the data of the user and different services to obtain the result data of the analysis and prediction,
the strategy management module carries out strategy configuration on the existing original data and the result data which is analyzed and predicted according to the service requirement, utilizes the rule engine to carry out one-by-one inspection on the result data of the strategy configuration, feeds back the result data of the strategy configuration to form different strategy execution results, distinguishes the user credit degree,
and the financial product supervision module is used for configuring corresponding financial products according to the user credit and supervising the user.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201911074312.6A CN110837963A (en) | 2019-11-06 | 2019-11-06 | Risk control platform construction method based on data, model and strategy |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201911074312.6A CN110837963A (en) | 2019-11-06 | 2019-11-06 | Risk control platform construction method based on data, model and strategy |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN110837963A true CN110837963A (en) | 2020-02-25 |
Family
ID=69576266
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201911074312.6A Withdrawn CN110837963A (en) | 2019-11-06 | 2019-11-06 | Risk control platform construction method based on data, model and strategy |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN110837963A (en) |
Cited By (13)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111353882A (en) * | 2020-04-17 | 2020-06-30 | 新分享科技服务(深圳)有限公司 | Privatized deployment retail asset wind control method and device and electronic equipment |
| CN111402048A (en) * | 2020-03-11 | 2020-07-10 | 杭州首新网络科技有限公司 | Intelligent wind control system capable of processing mass financial big data |
| CN111507820A (en) * | 2020-04-08 | 2020-08-07 | 上海金仕达软件科技有限公司 | Gridding financial wind control intelligent system |
| CN111640000A (en) * | 2020-04-17 | 2020-09-08 | 四川新网银行股份有限公司 | Data source calling method based on real-time decision |
| CN111898931A (en) * | 2020-08-24 | 2020-11-06 | 深圳市富之富信息科技有限公司 | Variable-based strategic wind control engine implementation method and device and computer equipment |
| CN111934793A (en) * | 2020-07-31 | 2020-11-13 | 中国工商银行股份有限公司 | Internet architecture full link monitoring method and device |
| CN112100602A (en) * | 2020-07-22 | 2020-12-18 | 武汉极意网络科技有限公司 | Strategy monitoring and optimizing system and method based on verification code product |
| CN113256404A (en) * | 2021-06-16 | 2021-08-13 | 浙江网商银行股份有限公司 | Data processing method and device |
| CN113610535A (en) * | 2021-07-29 | 2021-11-05 | 浙江惠瀜网络科技有限公司 | Risk monitoring method and device suitable for consumption staging business process |
| CN113628036A (en) * | 2021-08-16 | 2021-11-09 | 武汉众邦银行股份有限公司 | Big data risk detection model-based method and device |
| CN117710081A (en) * | 2023-11-29 | 2024-03-15 | 浙江孚临科技有限公司 | Information service processing system for financial risk control |
| CN117852926A (en) * | 2024-03-04 | 2024-04-09 | 四川享宇科技有限公司 | Champion challenger strategy management method and champion challenger strategy management system |
| CN117852926B (en) * | 2024-03-04 | 2024-05-14 | 四川享宇科技有限公司 | Champion challenger strategy management method and champion challenger strategy management system |
-
2019
- 2019-11-06 CN CN201911074312.6A patent/CN110837963A/en not_active Withdrawn
Cited By (15)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111402048A (en) * | 2020-03-11 | 2020-07-10 | 杭州首新网络科技有限公司 | Intelligent wind control system capable of processing mass financial big data |
| CN111507820A (en) * | 2020-04-08 | 2020-08-07 | 上海金仕达软件科技有限公司 | Gridding financial wind control intelligent system |
| CN111353882A (en) * | 2020-04-17 | 2020-06-30 | 新分享科技服务(深圳)有限公司 | Privatized deployment retail asset wind control method and device and electronic equipment |
| CN111640000A (en) * | 2020-04-17 | 2020-09-08 | 四川新网银行股份有限公司 | Data source calling method based on real-time decision |
| CN112100602A (en) * | 2020-07-22 | 2020-12-18 | 武汉极意网络科技有限公司 | Strategy monitoring and optimizing system and method based on verification code product |
| CN112100602B (en) * | 2020-07-22 | 2023-11-10 | 武汉极意网络科技有限公司 | Policy monitoring and optimizing system and method based on verification code product |
| CN111934793A (en) * | 2020-07-31 | 2020-11-13 | 中国工商银行股份有限公司 | Internet architecture full link monitoring method and device |
| CN111898931A (en) * | 2020-08-24 | 2020-11-06 | 深圳市富之富信息科技有限公司 | Variable-based strategic wind control engine implementation method and device and computer equipment |
| CN111898931B (en) * | 2020-08-24 | 2024-04-30 | 深圳市富之富信息科技有限公司 | Variable-based strategy type wind control engine implementation method and device and computer equipment |
| CN113256404A (en) * | 2021-06-16 | 2021-08-13 | 浙江网商银行股份有限公司 | Data processing method and device |
| CN113610535A (en) * | 2021-07-29 | 2021-11-05 | 浙江惠瀜网络科技有限公司 | Risk monitoring method and device suitable for consumption staging business process |
| CN113628036A (en) * | 2021-08-16 | 2021-11-09 | 武汉众邦银行股份有限公司 | Big data risk detection model-based method and device |
| CN117710081A (en) * | 2023-11-29 | 2024-03-15 | 浙江孚临科技有限公司 | Information service processing system for financial risk control |
| CN117852926A (en) * | 2024-03-04 | 2024-04-09 | 四川享宇科技有限公司 | Champion challenger strategy management method and champion challenger strategy management system |
| CN117852926B (en) * | 2024-03-04 | 2024-05-14 | 四川享宇科技有限公司 | Champion challenger strategy management method and champion challenger strategy management system |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110837963A (en) | Risk control platform construction method based on data, model and strategy | |
| US11941650B2 (en) | Explainable machine learning financial credit approval model for protected classes of borrowers | |
| CN110717828B (en) | Abnormal account detection method and system based on frequent transaction mode | |
| CN112766550B (en) | Random forest-based power failure sensitive user prediction method, system, storage medium and computer equipment | |
| CN110334737A (en) | A kind of method and system of the customer risk index screening based on random forest | |
| CN112668944A (en) | Enterprise wind control method, device, equipment and medium based on big data credit investigation | |
| CN112767136A (en) | Credit anti-fraud identification method, credit anti-fraud identification device, credit anti-fraud identification equipment and credit anti-fraud identification medium based on big data | |
| Li et al. | Transaction fraud detection using gru-centered sandwich-structured model | |
| CN113537807B (en) | Intelligent wind control method and equipment for enterprises | |
| CN111986027A (en) | Abnormal transaction processing method and device based on artificial intelligence | |
| CN110728301A (en) | Credit scoring method, device, terminal and storage medium for individual user | |
| Tsai et al. | Data pre-processing by genetic algorithms for bankruptcy prediction | |
| Yahaya et al. | An enhanced bank customers churn prediction model using a hybrid genetic algorithm and k-means filter and artificial neural network | |
| CN112686446A (en) | Machine learning interpretability-oriented credit default prediction method and system | |
| CN115115369A (en) | Data processing method, device, equipment and storage medium | |
| CN116091206B (en) | Credit evaluation method, credit evaluation device, electronic equipment and storage medium | |
| CN113052422A (en) | Wind control model training method and user credit evaluation method | |
| CN116342255A (en) | Internet consumption credit anti-fraud risk identification method and system | |
| CN115496364A (en) | Method and device for identifying heterogeneous enterprises, storage medium and electronic equipment | |
| CN114021612A (en) | Novel personal credit assessment method and system | |
| KR102499182B1 (en) | Loan regular auditing system using artificia intellicence | |
| CN111951099A (en) | Credit card issuing model and application method thereof | |
| CN117745080B (en) | Multi-factor authentication-based data access control and security supervision method and system | |
| Kong | Prediction of Enterprise Economy by Network Communication Business Based on BP Neural Network in Big Data Security Environment | |
| KR102499183B1 (en) | Loan regular auditing system using artificia intellicence |
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 | ||
| WW01 | Invention patent application withdrawn after publication |
Application publication date: 20200225 |
|
| WW01 | Invention patent application withdrawn after publication |