CN113393316B - Loan overall process accurate wind control and management system based on massive big data and core algorithm - Google Patents
Loan overall process accurate wind control and management system based on massive big data and core algorithm Download PDFInfo
- Publication number
- CN113393316B CN113393316B CN202110649992.0A CN202110649992A CN113393316B CN 113393316 B CN113393316 B CN 113393316B CN 202110649992 A CN202110649992 A CN 202110649992A CN 113393316 B CN113393316 B CN 113393316B
- Authority
- CN
- China
- Prior art keywords
- module
- loan
- risk
- data
- wind control
- 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.)
- Active
Links
- 238000004422 calculation algorithm Methods 0.000 title claims abstract description 80
- 238000000034 method Methods 0.000 title claims abstract description 77
- 230000008569 process Effects 0.000 title claims abstract description 60
- 238000012549 training Methods 0.000 claims abstract description 51
- 238000012544 monitoring process Methods 0.000 claims abstract description 37
- 238000012545 processing Methods 0.000 claims abstract description 20
- 238000011156 evaluation Methods 0.000 claims abstract description 12
- 230000006870 function Effects 0.000 claims abstract description 10
- 239000002245 particle Substances 0.000 claims description 51
- 238000007726 management method Methods 0.000 claims description 49
- 238000004458 analytical method Methods 0.000 claims description 22
- 238000012216 screening Methods 0.000 claims description 22
- 238000003860 storage Methods 0.000 claims description 22
- 238000012360 testing method Methods 0.000 claims description 21
- 238000004364 calculation method Methods 0.000 claims description 20
- 238000005457 optimization Methods 0.000 claims description 20
- 238000007619 statistical method Methods 0.000 claims description 18
- 238000013528 artificial neural network Methods 0.000 claims description 16
- 230000006399 behavior Effects 0.000 claims description 16
- 238000004088 simulation Methods 0.000 claims description 16
- 238000010801 machine learning Methods 0.000 claims description 13
- 238000005259 measurement Methods 0.000 claims description 13
- 238000012954 risk control Methods 0.000 claims description 13
- 238000012502 risk assessment Methods 0.000 claims description 12
- 238000007405 data analysis Methods 0.000 claims description 11
- 238000005516 engineering process Methods 0.000 claims description 5
- 238000000605 extraction Methods 0.000 claims description 5
- 238000013058 risk prediction model Methods 0.000 claims description 4
- 238000004140 cleaning Methods 0.000 claims description 3
- 238000004891 communication Methods 0.000 claims description 3
- 238000012805 post-processing Methods 0.000 claims description 3
- 238000012217 deletion Methods 0.000 claims description 2
- 230000037430 deletion Effects 0.000 claims description 2
- 230000014759 maintenance of location Effects 0.000 claims description 2
- 230000008030 elimination Effects 0.000 claims 1
- 238000003379 elimination reaction Methods 0.000 claims 1
- 238000013461 design Methods 0.000 abstract 1
- 238000004590 computer program Methods 0.000 description 8
- 230000006872 improvement Effects 0.000 description 8
- 238000010586 diagram Methods 0.000 description 6
- 239000011159 matrix material Substances 0.000 description 6
- 230000008520 organization Effects 0.000 description 6
- 230000007547 defect Effects 0.000 description 5
- 230000000875 corresponding effect Effects 0.000 description 3
- 239000006185 dispersion Substances 0.000 description 3
- 230000005540 biological transmission Effects 0.000 description 2
- 230000002596 correlated effect Effects 0.000 description 2
- 238000005336 cracking Methods 0.000 description 2
- 238000013523 data management Methods 0.000 description 2
- 230000000694 effects Effects 0.000 description 2
- 238000011835 investigation Methods 0.000 description 2
- 238000004900 laundering Methods 0.000 description 2
- 239000000463 material Substances 0.000 description 2
- 230000003287 optical effect Effects 0.000 description 2
- 230000001133 acceleration Effects 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000001276 controlling effect Effects 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 238000007689 inspection Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 230000003068 static effect Effects 0.000 description 1
- 230000000007 visual effect Effects 0.000 description 1
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
- G06Q40/00—Finance; Insurance; Tax strategies; Processing of corporate or income taxes
- G06Q40/03—Credit; Loans; Processing thereof
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
- G06N3/084—Backpropagation, e.g. using gradient descent
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Business, Economics & Management (AREA)
- General Physics & Mathematics (AREA)
- General Health & Medical Sciences (AREA)
- Accounting & Taxation (AREA)
- Software Systems (AREA)
- Mathematical Physics (AREA)
- General Engineering & Computer Science (AREA)
- Computing Systems (AREA)
- Molecular Biology (AREA)
- Finance (AREA)
- Health & Medical Sciences (AREA)
- Life Sciences & Earth Sciences (AREA)
- Artificial Intelligence (AREA)
- Biomedical Technology (AREA)
- Biophysics (AREA)
- Computational Linguistics (AREA)
- Data Mining & Analysis (AREA)
- Evolutionary Computation (AREA)
- Marketing (AREA)
- Development Economics (AREA)
- General Business, Economics & Management (AREA)
- Economics (AREA)
- Technology Law (AREA)
- Strategic Management (AREA)
- Financial Or Insurance-Related Operations Such As Payment And Settlement (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention relates to the technical field of wind control management, in particular to a loan overall-process accurate wind control and management system based on massive big data and a core algorithm. The system comprises a data processing unit, a training model unit, a safety management unit and a decision system unit; the data processing unit is used for acquiring and processing massive data information; the training model unit is used for building a prediction recognition model related to loan wind control and carrying out training and learning; the safety management unit is used for carrying out risk monitoring and safety management on the whole loan service process; and the decision system unit is used for taking the risk factors as the basis of wind control decision. The design of the invention can rapidly evaluate the credit conditions of the loan users and the loan placement institutions by analyzing the data; the system can quickly identify the risk condition of loan and provide functions of risk identification evaluation, post-loan early warning and the like; the method can optimize the wind control process and make corresponding decision adjustment, so that risks can be quickly and accurately found and controlled, and the workload and difficulty of loan approval are reduced.
Description
Technical Field
The invention relates to the technical field of wind control management, in particular to a loan overall-process accurate wind control and management system based on massive big data and a core algorithm.
Background
The loan wind control is risk control, and means that a risk manager takes various measures and methods to eliminate or reduce various possibilities of occurrence of a risk event, or a risk controller reduces losses caused when the risk event occurs. The premise of risk control is to discover risks, and the possible risks in the loan service mainly include three types, namely credit risk, transaction risk and account wind. With the rapid development and expansion of financial industry, financial information of loan users is generally scattered in different financial institutions, it is difficult to quickly collect past credit information of users in a loan business process, loan risks are difficult to quickly identify due to the defect of information isolated islands, the workload and difficulty of credit approval are increased, risk control cannot be quickly and accurately performed, and further economic loss of loan institutions due to the fact that wind control is not in place is possible.
Disclosure of Invention
The invention aims to provide a loan overall process accurate wind control and management system based on massive big data and a core algorithm, so as to solve the problems in the background technology.
In order to solve the technical problems, one of the objectives of the present invention is to provide a loan process accurate wind control and management system based on massive big data and core algorithm, which comprises
The system comprises a data processing unit, a training model unit, a safety management unit and a decision system unit; the signal output end of the data processing unit is connected with the signal input end of the training model unit, the signal output end of the training model unit is connected with the signal input end of the safety management unit, and the signal output end of the safety management unit is connected with the signal input end of the decision system unit; the data processing unit is used for acquiring massive data information related to the loan through a multi-dimensional data acquisition way and carrying out duplication removal, storage, statistical analysis and division processing on the data; the training model unit is used for constructing a prediction recognition model related to loan wind control based on BP neural network prediction technology and carrying out training and learning; the safety management unit is used for carrying out risk monitoring and safety management on the whole process of the loan service; the decision system unit is used for taking risk factors possibly existing in the loan business predicted according to the model as the basis of wind control decision;
the data processing unit comprises an acquisition and duplication removal module, a classification storage module, a statistical analysis module and a data set division module;
the training model unit comprises a model building module, an algorithm training module, a relevant inspection module and a machine learning module;
the safety management unit comprises a behavior identification module, a credit rating module, a risk evaluation module and a post-credit early warning module;
the decision system unit comprises an index calculation module, an operation auditing module, a wind control decision module and a comprehensive report module.
As a further improvement of the technical solution, a signal output end of the collection and duplicate removal module is connected to a signal input end of the classification storage module, a signal output end of the classification storage module is connected to a signal input end of the statistical analysis module, and a signal output end of the statistical analysis module is connected to a signal input end of the number set dividing module; the collection and duplication removal module is used for obtaining data of users and institutions related to the loan service from stock data inside an enterprise, an online cloud database and a third-party reliable data platform and carrying out duplication removal and cleaning operation on the data; the classified storage module is used for classifying, summarizing and storing mass data in a distributed manner; the statistical analysis module is used for counting and analyzing the related data; the number set dividing module is used for randomly extracting partial data from the database according to the requirement of model training and dividing the partial data into a training number set and a testing number set according to a certain proportion.
The sources of the data comprise an information management system in a loan institution, an information management system of a cooperative bank, reliable public data on line, a big data transaction center, a Chinese people's bank and the like.
As a further improvement of the technical scheme, a signal output end of the model building module is connected with a signal input end of the algorithm training module, a signal output end of the algorithm training module is connected with a signal input end of the related checking module, and a signal output end of the related checking module is connected with a signal input end of the machine learning module; the model building module is used for building prediction models of various risk types in the loan on the basis of a BP neural network algorithm; the algorithm training module is used for carrying out algorithm training on the prediction model according to certain flow steps; the correlation testing module is used for analyzing and testing the correlation between the prediction results of various risk types through a Pearson correlation testing algorithm and a Spanish correlation testing algorithm respectively; the machine learning module is used for perfecting the prediction algorithm through machine learning so as to improve the accuracy of prediction.
As a further improvement of the technical scheme, the model building module comprises a credit evaluation module, a transaction monitoring module, an account security module and a risk control module; the credit evaluation module, the transaction monitoring module, the account security module and the risk control module operate in parallel; the credit evaluation module is used for building a credit evaluation prediction model to predict the credit conditions of the user and the loan institution; the transaction monitoring module is used for building a transaction monitoring prediction model to predict the transaction process and the existing risk in the user loan service process; the account security module is used for predicting the bank account information and related risks applied to the loan service; the risk control module is used for continuously optimizing and iterating each risk prediction model so as to more quickly identify and manage all risk types possibly existing in the loan business process.
The credit risk comprises malicious overdue, deception, false identity, intermediary agency, cross-platform loan and the like; the transaction risk comprises card stealing, card embezzlement, bill swiping, cash register, false transaction, money laundering and the like; the account risk comprises malicious registration, account embezzlement, library dragging and crashing, brute force cracking, account attack and the like.
As a further improvement of the technical solution, the flow of the algorithm training module includes the following steps:
step1, forward transmission phase:
1.4 taking a sample P from the set of samples i ，Q i From P to P i Inputting a network;
1.5, calculating error measure E 1 And an actual output O i ＝F L (...(F 2 (F 1 (P i W (1) )W (2) )...)W (L) )；
1.6, pair weight value W (1) ，W (2) ，...，W (L) Each making an adjustment, repeating the cycle until E i ＜ε；
Step2, backward propagation stage — error propagation stage:
2.1 calculating the actual output O P And an ideal output Q i A difference of (d);
2.2, adjusting the weight matrix of the output layer by using the error of the output layer;
2.3 calculating Global error
2.4, estimating the error of a direct front layer of the output layer by using the error, and estimating the error of a previous layer by using the error of the front layer of the output layer, thereby obtaining the error estimation of all other layers;
2.5, modifying the weight matrix by using the estimation to form a process of gradually transmitting the error shown by the output end to the output end along the direction opposite to the output signal;
in addition, the error measure calculation formula of the network with respect to the whole sample set is: e = ∑ Σ i E i 。
As a further improvement of the present technical solution, the calculation expression of the correlation check module is as follows:
pearson correlation coefficient test algorithm:
wherein,
the Spireman correlation coefficient test algorithm:
wherein d is i ＝X i -Y i ，d i Is the difference in level, p s Is of the value [ -1, +1]；
Furthermore, the pearson algorithm is used for continuous, normally distributed and linearly correlated data, and the spearman algorithm is used for any data that does not satisfy the above conditions.
As a further improvement of the technical scheme, the behavior identification module, the credit rating module, the risk assessment module and the post-credit warning module are sequentially connected through ethernet communication and run in parallel; the behavior recognition module is used for monitoring the behaviors of the user and the institution in the loan service process and recognizing the behavior with risk; the credit rating module is used for rating the credit conditions of the users and the organizations by combining big data analysis and prediction of the training model; the risk evaluation module is used for carrying out omnibearing analysis and judgment on the risk and the risk condition possibly existing in the whole process of the loan service; the post-loan early warning module is used for carrying out early warning analysis on the possible risk condition in the post-processing process of the loan business according to the risk assessment condition of multiple parties.
As a further improvement of the technical scheme, the risk assessment module comprises a risk simulation module, a measurement early warning module, a dynamic monitoring module and a process optimization module; the signal output end of the risk simulation module is connected with the signal input end of the measurement early warning module, the signal output end of the measurement early warning module is connected with the signal input end of the dynamic monitoring module, and the signal output end of the dynamic monitoring module is connected with the signal input end of the flow optimization module; the risk simulation module is used for carrying out simulation analysis on the possible risks through an algorithm model; the measurement early warning module is used for measuring the risk and the risk size obtained through simulation analysis and sending early warning to related personnel when the risk size reaches a certain degree; the dynamic monitoring module is used for dynamically monitoring risks in the whole process of the loan service based on big data analysis; and the flow optimization module is used for optimizing and adjusting the business flow with the risk according to the risk condition of the comprehensive analysis.
As a further improvement of the technical solution, a signal output end of the index calculation module is connected to a signal input end of the operation auditing module, a signal output end of the operation auditing module is connected to a signal input end of the wind control decision module, and a signal output end of the wind control decision module is connected to a signal input end of the comprehensive report module; the index calculation module is used for carrying out real-time flow calculation and graph calculation on the loan service related data according to a preset index center; the operation auditing module is used for auditing and auditing the loan business case on the operation auditing platform in a random extraction mode; the wind control decision module is used for integrating the result analysis of the rule decision and the model decision to select and implement the risk decision; the comprehensive report module is used for collating the data of the loan service related to the full process such as data analysis, risk prediction, operation auditing, wind control decision making and the like to form a corresponding report.
The invention also aims to provide an operation method of the loan overall process accurate wind control and management system based on massive big data and a core algorithm, which comprises the following steps:
s1, a user submits a loan application in a loan institution, and the institution automatically forms a loan service;
s2, the system platform automatically acquires stock data from an internal information management system of the organization, and simultaneously acquires data information related to the user and the organization from a three-party reliable transaction data management platform;
s3, carrying out statistical analysis on the data to obtain the credit conditions of the user and the organization;
s4, importing the data into a pre-built and trained model, and obtaining the predicted risk type and risk size results;
s5, submitting loan service, and auditing the service case by operating an auditing platform, wherein the auditing process is performed by random extraction so as to reduce the risk of malicious operation;
s6, combining data analysis, prediction identification and case auditing results, analyzing and predicting risks possibly existing in the whole loan service process, early warning the risks, and making a wind control decision according to the risk condition;
and S7, forming a wind control report, submitting the wind control analysis report together with the loan materials as a basis for loan approval, reducing the workload of loan risk investigation, and reducing the risk of loan business.
The invention also aims to provide an operation device of the loan overall process accurate wind control and management system based on the massive large data and the core algorithm, which comprises a processor, a memory and a computer program stored in the memory and operated on the processor, wherein the processor is used for realizing any loan overall process accurate wind control and management system based on the massive large data and the core algorithm when executing the computer program.
The present invention also provides a computer-readable storage medium, wherein a computer program is stored in the computer-readable storage medium, and when the computer program is executed by a processor, the system realizes the above-mentioned loan overall process accurate pneumatic control and management system based on massive big data and core algorithm.
Compared with the prior art, the invention has the beneficial effects that:
1. the loan overall process accurate wind control and management system based on massive big data and a core algorithm acquires massive multi-party data in a multi-dimensional data acquisition mode, eliminates information isolated islands, and can quickly evaluate the credit conditions of loan users and loan-putting institutions by analyzing the data;
2. the loan overall-process accurate wind control and management system based on massive big data and a core algorithm is used for rapidly identifying the loan risk condition through a risk prediction model established by a BP neural network prediction technology and training and learning, and providing functions of risk identification evaluation, post-loan early warning and the like;
3. the loan overall-process accurate wind control and management system based on massive big data and a core algorithm can be combined with a dynamic risk monitoring result, optimize a wind control process and make corresponding decision adjustment, so that risks can be quickly and accurately found and controlled, workload and difficulty of loan approval are reduced, possibility of occurrence of risk events is reduced, or economic loss possibly caused when the risk events occur is reduced.
4. The loan overall process accurate wind control and management system based on massive big data and a core algorithm adopts a BP neural network algorithm to carry out algorithm training on a prediction model, adjusts the weight matrix of an output layer by using the error of the output layer, estimates the error of a direct leading layer of the output layer by using the error, and estimates the error of a more previous layer by using the error of the leading layer of the output layer. Error estimates for all other layers are thus obtained and used to implement the modification of the weight matrix. A process is formed to pass the error exhibited by the output terminal in a stepwise manner in the opposite direction to the input signal to the input terminal. Secondly, a Pearson algorithm is adopted for correlation coefficient test, and different from the traditional mode, the standard deviation mode is adopted, the standard deviation and the variable calculation unit are the same, the ratio difference is clear, and the covariance expresses the expectation of the total error of the two variables in visual sense. If the variation trends of the two variables are consistent, namely if one of the two variables is greater than the expected value of the other variable is also greater than the expected value of the other variable, the covariance between the two variables is a positive value; if the two variables have opposite trend, i.e. one variable is larger than the expected value but the other variable is smaller than the expected value, the covariance between the two variables is negative.
Drawings
FIG. 1 is a block diagram of an exemplary product architecture of the present invention;
FIG. 2 is a block diagram of the overall system apparatus of the present invention;
FIG. 3 is a diagram of one embodiment of a local system device architecture;
FIG. 4 is a second block diagram of a local system apparatus according to the present invention;
FIG. 5 is a third block diagram of a local system apparatus according to the present invention;
FIG. 6 is a fourth embodiment of the present invention;
FIG. 7 is a fifth embodiment of the present invention;
FIG. 8 is a fifth embodiment of the present invention;
FIG. 9 is a block diagram of an exemplary electronic computer product device of the present invention.
The various reference numbers in the figures mean:
100. a data processing unit; 101. a collection duplication removal module; 102. a classification storage module; 103. a statistical analysis module; 104. a number set dividing module;
200. training a model unit; 201. a model building module; 2011. a credit assessment module; 2012. a transaction monitoring module; 2013. an account security module; 2014. a risk control module; 202. an algorithm training module; 203. a correlation check module; 204. a machine learning module;
300. a security management unit; 301. a behavior recognition module; 302. a credit rating module; 303. a risk assessment module; 3031. a risk simulation module; 3032. a measurement early warning module; 3033. a dynamic monitoring module; 3034. a flow optimization module; 304. a post-loan warning module;
400. a decision system unit; 401. an index calculation module; 402. an operation auditing module; 403. a wind control decision module; 404. and a comprehensive report module.
Detailed Description
The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention, and it is obvious that the described embodiments are only a part of the embodiments of the present invention, and not all of the embodiments. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.
Example 1
As shown in fig. 1-9, the present embodiment provides a loan overall process accurate wind control and management system based on massive big data and core algorithm, which includes
A data processing unit 100, a training model unit 200, a security management unit 300 and a decision system unit 400; the signal output end of the data processing unit 100 is connected with the signal input end of the training model unit 200, the signal output end of the training model unit 200 is connected with the signal input end of the security management unit 300, and the signal output end of the security management unit 300 is connected with the signal input end of the decision system unit 400; the data processing unit 100 is used for acquiring massive data information related to the loan through a multi-dimensional data acquisition way and performing duplication removal, storage, statistical analysis and division processing on the data; the training model unit 200 is used for building a prediction recognition model related to loan wind control based on BP neural network prediction technology and carrying out training and learning; the safety management unit 300 is used for performing risk monitoring and safety management on the whole loan service process; the decision system unit 400 is used for predicting risk factors possibly existing in the loan service according to the model to be used as the basis of wind control decision;
the data processing unit 100 comprises an acquisition and de-weighting module 101, a classification storage module 102, a statistical analysis module 103 and a dataset division module 104;
the training model unit 200 comprises a model building module 201, an algorithm training module 202, a correlation testing module 203 and a machine learning module 204;
the security management unit 300 comprises a behavior recognition module 301, a credit rating module 302, a risk assessment module 303 and a post-credit warning module 304;
the decision system unit 400 includes an index calculation module 401, an operation auditing module 402, a wind control decision module 403, and a comprehensive reporting module 404.
In this embodiment, the signal output end of the acquisition and deduplication module 101 is connected to the signal input end of the classification storage module 102, the signal output end of the classification storage module 102 is connected to the signal input end of the statistical analysis module 103, and the signal output end of the statistical analysis module 103 is connected to the signal input end of the number set dividing module 104; the collection duplication removal module 101 is used for obtaining data of users and institutions related to loan service from stock data inside an enterprise, an online cloud database and a third-party reliable data platform and performing duplication removal and cleaning operation on the data; the classification storage module 102 is used for classifying, summarizing and storing mass data in a distributed manner; the statistical analysis module 103 is used for performing statistics and analysis on the related data; the number set dividing module 104 is configured to randomly extract a part of data from the database according to the requirement of model training and divide the part of data into a training number set and a testing number set according to a certain proportion.
The data sources include information management systems in loan institutions, information management systems of cooperative banks, reliable public data on the internet, big data transaction centers, china people's banks and the like.
In this embodiment, the signal output end of the model building module 201 is connected with the signal input end of the algorithm training module 202, the signal output end of the algorithm training module 202 is connected with the signal input end of the relevant checking module 203, and the signal output end of the relevant checking module 203 is connected with the signal input end of the machine learning module 204; the model building module 201 is used for building prediction models of various risk types in the loan on the basis of a BP neural network algorithm; the algorithm training module 202 is used for performing algorithm training on the prediction model according to certain process steps; the correlation testing module 203 is used for analyzing and testing the correlation between the prediction results of various risk types through a Pearson correlation testing algorithm and a Spanish correlation testing algorithm respectively; the machine learning module 204 is used to refine the prediction algorithm by machine learning to improve the accuracy of the prediction.
Further, the model building module 201 includes a credit assessment module 2011, a transaction monitoring module 2012, an account security module 2013 and a risk control module 2014; the credit rating module 2011, the transaction monitoring module 2012, the account security module 2013, and the risk control module 2014 operate in parallel; the credit assessment module 2011 is used for building a credit assessment prediction model to predict the credit condition of the user and the lending institution; the transaction monitoring module 2012 is used for building a transaction monitoring prediction model to predict the transaction process and the existing risk in the user loan business process; the account security module 2013 is used for predicting bank account information and related risks for applying to loan services; the risk control module 2014 is used to continuously optimize and iterate the various risk prediction models to more quickly identify and manage all the risk types that may exist in the loan transaction flow.
The credit risk comprises malicious overdue, deception loan, false identity, intermediary agency, cross-platform loan and the like; the transaction risk comprises card stealing, card embezzlement, bill swiping, cash register, false transaction, money laundering and the like; the account risk comprises malicious registration, account embezzlement, library dragging and crashing, brute force cracking, account attack and the like.
Preferably, the initial weight and the threshold of the BP neural network adopted in the process of predicting the model by the model building module (201) are optimized by a particle swarm algorithm, and the smaller the fitness function value of the current solution of the particles in the particle swarm is, the better the optimization effect of the particles is. The initial weight and the threshold of the BP neural network are optimized through the improved particle swarm algorithm, the improved particle swarm algorithm can effectively avoid being involved in local optimization, so that the defects that the BP neural network is low in convergence speed and easy to be involved in local optimization can be effectively overcome, and the optimized BP neural network has high evaluation accuracy.
Preferably, in the particle swarm algorithm, let
Represents the solution of the particles i in the population after the t iteration updating,
for the locally better solution, qbest, for learning of particle i in the defined population at the (t + 1) th iterative update t Representing the global optimal solution of the population after the t iterative update, and leading the particles i in the population to pass through a local better solution
And the global optimal solution Qbest t Learning is performed to obtain the solution of the particle i after the (t + 1) th iteration update
The local preferred solution
The values of (A) are:
is provided with
Represents the current historical optimal solution superior to the particle i in the population after the t iteration update
Is composed of a set of solutions, and
wherein,
representation collection
The j-th better solution in (a),
representation collection
The number of better solutions in (a) is,
indicates a better solution
The value of the fitness function of (a) is,
representing the historical optimal solution of the particles i in the population after the updating of the t iteration,
representing historical optimal solutions
The fitness function value of (a);
to the collection
The better solutions in (1) are sequentially screened and defined
Representing for better solution
The set individual screening coefficient is set according to the specific screening method,
representing a target set
Overall screening coefficient set by medium-preferred solution, and
and
the values of (A) are respectively:
in the formula,
indicates a better solution
The optimizing screening factor of, and
indicates a better solution
A distance screening factor of, and
indicates a better solution
Iteratively adjust the factor, and
where T denotes the current number of iterations, T max The maximum number of iterations is indicated,
and
the effect of (a) is to take the maximum value,
the function of (2) is to take the median value;
when the collection
The better solution of
Satisfies the following conditions:
then in the set
Middle retention better solution
But rather the solution is better
Satisfies the following conditions:
then in the set
Better solution of medium deletion
Is provided with
Set of representation pairs
The better solutions in the above are sequentially screened, and the local better solutions are selected from the group consisting of the rest better solutions
I.e. in the aggregate
One of the preferred solutions is selected randomly.
Preferably, the particles i in the population are better solved locally by
And the global optimal solution Qbest t Learning is performed to obtain the solution of the particle i after the (t + 1) th iteration update
The method comprises the following specific steps:
wherein,
represents the step size of the particle i in the population at the (t + 1) th iteration update,
representing the step size of the particle i in the population at the time of the t-th iterative update, c 1 And c 2 Learning factor, r, representing a population 1 And r 2 Respectively representThe machine generates a random number between 0 and 1, ω represents the inertial weight factor of the particles in the population.
The optimal particle swarm optimization method comprises the steps that an iterative updating process of a traditional particle swarm algorithm is improved, the traditional particle swarm algorithm has the defects that the convergence speed is low and the probability of falling into local optimum is high, therefore, the traditional particle swarm algorithm is used for optimizing an initial weight and a threshold of a BP neural network, the defects that the convergence speed of the BP neural network is low and the probability of falling into local extreme is not effectively solved, and the traditional particle swarm algorithm needs to be improved if the defects that the convergence speed of the BP neural network is low and the probability of falling into extreme is really solved through the particle swarm algorithm; the local better solutions of the particles are randomly selected from historical better solutions which are superior to the particles in a population, an individual screening coefficient and an overall screening coefficient are set for the better solutions, optimization screening factors in the individual screening coefficients are used for guaranteeing forward convergence of the particles, distance screening factors in the individual screening coefficients are used for guaranteeing local optimization of the particles, the overall screening coefficient controls the number of the rest better solutions in a set through the dispersion degree of the individual screening coefficients between the better solutions in the population, when the dispersion degree of the individual screening coefficients between the better solutions is small, the better solutions in the population and the particles are distributed more uniformly, at the moment, the value of the overall screening coefficient is small, so that more better solutions are reserved in a better solution set, the diversity of the population can be increased when the local better solutions are selected, the capacity of the local optimal is increased, when the value of the individual screening coefficients between the better solutions is large, the dispersion of the better solutions and the particles distributed between the particles in the population is indicated, and when the overall screening coefficient is large, the convergence rate of the particle population is increased, the better solutions are reserved in the particle population, and the acceleration speed of the better solutions is low. That is, in the preferred embodiment, the particle swarm is made to adaptively adjust the particle optimization mode according to the current optimization situation by learning the particle to the local better solution, so that the convergence speed of the particle swarm is accelerated, and the capability of the particle swarm to jump out of the local optimal solution is enhanced.
Specifically, the flow of the algorithm training module 202 includes the following steps:
step1, forward transmission phase:
1.7 taking a sample P from the set of samples i ，Q i A 1 is to P i Inputting a network;
1.8, calculating the error measure E 1 And an actual output O i ＝F L (...(F 2 (F 1 (P i W (1) )W (2) )...)W (L) )；
1.9, pair weight value W (1) ，W (2) ，...，W (L) Each making an adjustment, repeating the cycle until Σ E i ＜ε；
Step2, backward propagation stage-error propagation stage:
2.1 calculating the actual output O P And an ideal output Q i A difference of (d);
2.2, adjusting the weight matrix of the output layer by using the error of the output layer;
2.3 calculating Global error
2.4, estimating the error of a direct leading layer of the output layer by using the error, and estimating the error of a previous layer by using the error of the leading layer of the output layer, thereby obtaining the error estimation of all other layers;
2.5, modifying the weight matrix by using the estimation to form a process of gradually transmitting the error shown by the output end to the output end along the direction opposite to the output signal;
in addition, the error measure calculation formula of the network with respect to the whole sample set is: e = ∑ Σ i E i 。
Specifically, the computational expression of the correlation check module 203 is as follows:
pearson correlation coefficient test algorithm:
wherein,
the Spanish correlation coefficient checking algorithm:
wherein d is i ＝X i -Y i ，d i As difference in grade, p s Is of value [ -1, + 1)]；
Furthermore, the pearson algorithm is used for continuous, normally distributed and linearly correlated data, and the spearman algorithm is used for any data that does not satisfy the above conditions.
In this embodiment, the behavior recognition module 301, the credit rating module 302, the risk assessment module 303, and the post-credit warning module 304 are sequentially connected through ethernet communication and operate in parallel; the behavior identification module 301 is used for monitoring the behaviors of the user and the institution in the loan service process and identifying the behavior with risk; the credit rating module 302 is used for rating the credit conditions of the users and organizations by combining big data analysis and prediction of a training model; the risk evaluation module 303 is used for performing comprehensive analysis and judgment on risks and risk conditions which may exist in the whole loan business process; the post-loan early warning module 304 is used for performing early warning analysis on the risk condition which may exist in the post-processing process of the loan transaction according to the risk assessment condition of multiple parties.
Further, the risk assessment module 303 includes a risk simulation module 3031, a measurement early warning module 3032, a dynamic monitoring module 3033, and a process optimization module 3034; the signal output end of the risk simulation module 3031 is connected with the signal input end of the measurement early warning module 3032, the signal output end of the measurement early warning module 3032 is connected with the signal input end of the dynamic monitoring module 3033, and the signal output end of the dynamic monitoring module 3033 is connected with the signal input end of the flow optimization module 3034; the risk simulation module 3031 is used for performing simulation analysis on the possible risks through an algorithm model; the measurement early warning module 3032 is used for measuring the risk and the risk size obtained through simulation analysis and giving an early warning to related personnel when the risk size reaches a certain degree; the dynamic monitoring module 3033 is used for dynamically monitoring the risk in the whole process of the loan service based on big data analysis; the process optimization module 3034 is configured to perform optimization and adjustment on the business process with risk according to the risk condition of the comprehensive analysis.
In this embodiment, the signal output end of the index calculation module 401 is connected to the signal input end of the operation auditing module 402, the signal output end of the operation auditing module 402 is connected to the signal input end of the wind control decision module 403, and the signal output end of the wind control decision module 403 is connected to the signal input end of the comprehensive report module 404; the index calculation module 401 is used for performing real-time flow calculation and graph calculation on the loan service related data according to a preset index center; the operation auditing module 402 is used for auditing and auditing the loan transaction case on an operation auditing platform in a random extraction mode; the wind control decision module 403 is used for selecting and implementing risk decisions by integrating rule decisions and result analysis of model decisions; the comprehensive report module 404 is configured to collate the full-process data related to the loan service, such as data analysis, risk prediction, operation auditing, and wind control decision, and form a corresponding report.
The embodiment also provides an operation method of the loan overall process accurate wind control and management system based on massive big data and a core algorithm, which comprises the following steps:
s1, a user submits a loan application in a loan institution, and the institution automatically forms a loan service;
s2, the system platform automatically acquires stock data from an internal information management system of the organization, and simultaneously acquires data information related to the user and the organization from a three-party reliable transaction data management platform;
s3, carrying out statistical analysis on the data to obtain the credit conditions of the user and the organization;
s4, importing the data into a pre-built and trained model, and obtaining the predicted risk type and risk size results;
s5, submitting loan service, and auditing the service case by operating an auditing platform, wherein the auditing process is performed by random extraction so as to reduce the risk of malicious operation;
s6, combining data analysis, prediction identification and case auditing results, analyzing and predicting risks possibly existing in the whole loan service process, early warning the risks, and making a wind control decision according to the risk condition;
and S7, forming a wind control report, submitting the wind control analysis report together with the loan materials as a basis for loan approval, reducing the workload of loan risk investigation, and reducing the risk of loan business.
As shown in fig. 9, the embodiment further provides an operating device of the loan completion accurate pneumatic control and management system based on massive large data and core algorithm, and the operating device comprises a processor, a memory and a computer program stored in the memory and operated on the processor.
The processor comprises one or more than one processing core, the processor is connected with the processor through a bus, the memory is used for storing program instructions, and the loan overall process accurate pneumatic control and management system based on massive big data and a core algorithm is realized when the processor executes the program instructions in the memory.
Alternatively, the memory may be implemented by any type or combination of volatile or non-volatile memory devices, such as Static Random Access Memory (SRAM), electrically erasable programmable read-only memory (EEPROM), erasable programmable read-only memory (EPROM), programmable read-only memory (PROM), read-only memory (ROM), magnetic memory, flash memory, magnetic or optical disks.
In addition, the invention also provides a computer readable storage medium, wherein a computer program is stored in the computer readable storage medium, and when the computer program is executed by a processor, the system for accurately controlling and managing the whole loan process based on the massive big data and the core algorithm is realized.
Optionally, the present invention further provides a computer program product containing instructions, which when run on a computer, causes the computer to execute the above-mentioned aspects of the loan process accurate wind control and management system based on massive big data and core algorithm.
It will be understood by those skilled in the art that all or part of the steps for implementing the above embodiments may be implemented by hardware, or may be implemented by a program instructing relevant hardware, and the program may be stored in a computer-readable storage medium, where the above-mentioned storage medium may be a read-only memory, a magnetic disk or an optical disk, etc.
The foregoing shows and describes the general principles, principal features, and advantages of the invention. It will be understood by those skilled in the art that the present invention is not limited to the embodiments described above, and the preferred embodiments of the present invention are described in the above embodiments and the description, and are not intended to limit the present invention. The scope of the invention is defined by the appended claims and equivalents thereof.
Claims (6)
1. A loan overall process accurate wind control and management system based on massive big data and a core algorithm is characterized in that: the system comprises a data processing unit (100), a training model unit (200), a safety management unit (300) and a decision system unit (400); the signal output end of the data processing unit (100) is connected with the signal input end of the training model unit (200), the signal output end of the training model unit (200) is connected with the signal input end of the safety management unit (300), and the signal output end of the safety management unit (300) is connected with the signal input end of the decision system unit (400); the data processing unit (100) is used for acquiring massive data information related to loan through a multi-dimensional data acquisition way and performing duplication removal, storage, statistical analysis and division processing on the data; the training model unit (200) is used for building a prediction recognition model related to loan wind control based on BP neural network prediction technology and carrying out training and learning; the safety management unit (300) is used for carrying out risk monitoring and safety management on the whole loan service process; the decision system unit (400) is used for predicting risk factors existing in the loan service according to a model to be used as the basis of wind control decision;
the data processing unit (100) comprises an acquisition and de-duplication module (101), a classification storage module (102), a statistical analysis module (103) and a dataset division module (104);
the training model unit (200) comprises a model building module (201), an algorithm training module (202), a correlation testing module (203) and a machine learning module (204);
the security management unit (300) comprises a behavior recognition module (301), a credit rating module (302), a risk assessment module (303) and a post-credit warning module (304);
the decision system unit (400) comprises an index calculation module (401), an operation auditing module (402), a wind control decision module (403) and a comprehensive report module (404);
the signal output end of the model building module (201) is connected with the signal input end of the algorithm training module (202), the signal output end of the algorithm training module (202) is connected with the signal input end of the related checking module (203), and the signal output end of the related checking module (203) is connected with the signal input end of the machine learning module (204); the model building module (201) is used for building prediction models of various risk types in the loan on the basis of a BP neural network algorithm; the algorithm training module (202) is used for carrying out algorithm training on the prediction model according to the flow steps; the correlation testing module (203) is used for analyzing and testing the correlation between the prediction results of various risk types through a Pearson correlation testing algorithm and a Spanish correlation testing algorithm respectively; the machine learning module (204) is used for perfecting a prediction algorithm through machine learning so as to improve the accuracy of prediction;
optimizing initial weight and threshold of a BP neural network adopted in the process of predicting the model by the model building module (201) through a particle swarm algorithm; in the particle swarm optimization, let
Represents the solution of the particles i in the population after the t iteration updating,
for the locally better solution, qbest, for learning of particle i in the defined population at the (t + 1) th iterative update t Representing the global optimal solution of the population after the t-th iterative update, and leading the particles i in the population to pass through the local optimal solution
And the global optimal solution Qbest t Learning is performed to obtain the solution of the particle i after the (t + 1) th iteration update
The local preferred solution
The solving method comprises the following steps:
is provided with
Represents the current historical optimal solution superior to the particle i in the population after the t iteration update
Is composed of a set of solutions, and
wherein,
representation collection
The j-th better solution of (a) is,
representation collection
The number of better solutions in (a) is,
indicates a better solution
The value of the fitness function of (a),
representing the historical optimal solution of the particles i in the population after the t iteration update,
representing historical optimal solutions
A fitness function value of;
pair set
The better solutions in (1) are sequentially screened and defined
Represent for better solution
The set individual screening coefficient is set according to the specific screening method,
representing a target set
Overall screening coefficient set by medium and better solution, and
and
the values of (A) are respectively:
in the formula,
indicates a better solution
The optimizing screening factor of, and
indicates a better solution
A distance screening factor of, and
indicates a better solution
Iteratively adjust the factor, and
where T denotes the current number of iterations, T max The maximum number of iterations is indicated,
and
the function of (a) is to take the maximum value,
the function of (1) is to take the median value;
when the collection
The better solution of
Satisfies the following conditions:
then in the set
Middle retention better solution
While the best solution is
Satisfies the following conditions:
then in the set
Better solution of medium deletion
Is provided with
Set of representation pairs
The better solutions in the above steps are sequentially screened, and the local better solutions are selected from the set consisting of the rest better solutions
I.e. in the aggregate
Randomly selecting a better solution from the two solutions;
the particles i in the population are better solved locally
And the global optimal solution Qbest t The learning is carried out, and the learning is carried out,thereby obtaining the solution of the particle i after the (t + 1) th iteration update
The method specifically comprises the following steps:
wherein,
represents the step size of the particle i in the population at the (t + 1) th iteration update,
representing the step size of the particle i in the population at the time of the t-th iterative update, c 1 And c 2 Learning factor, r, representing a population 1 And r 2 Respectively, represent randomly generated random numbers between 0 and 1, and ω represents an inertial weight factor for the particles in the population.
2. The loan overall process accurate wind control and management system based on massive big data and core algorithm according to claim 1, characterized in that: the signal output end of the collection duplication elimination module (101) is connected with the signal input end of the classification storage module (102), the signal output end of the classification storage module (102) is connected with the signal input end of the statistical analysis module (103), and the signal output end of the statistical analysis module (103) is connected with the signal input end of the number set division module (104); the collection and duplicate removal module (101) is used for obtaining data of users and institutions related to the loan service from stock data inside an enterprise, an online cloud database and a third-party reliable data platform and performing duplicate removal and cleaning operation on the data; the classification storage module (102) is used for classifying, summarizing and storing massive data in a distributed manner; the statistical analysis module (103) is used for carrying out statistics and analysis on related data; the number set dividing module (104) is used for randomly extracting partial data from the database according to the requirement of model training and dividing the partial data into a training number set and a testing number set in proportion.
3. The loan overall process accurate wind control and management system based on massive big data and core algorithm according to claim 2, characterized in that: the model building module (201) comprises a credit evaluation module (2011), a transaction monitoring module (2012), an account security module (2013) and a risk control module (2014); the credit rating module (2011), the transaction monitoring module (2012), the account security module (2013), and the risk control module (2014) operate in parallel; the credit assessment module (2011) is used for building a credit assessment prediction model to predict the credit conditions of the user and the loan institution; the transaction monitoring module (2012) is used for constructing a transaction monitoring prediction model to predict the transaction process and the existing risk in the user loan business process; the account security module (2013) is used for predicting bank account information and related risks applied to loan services; the risk control module (2014) is used for continuously optimizing and iterating each risk prediction model so as to more quickly identify and manage all risk types existing in the loan business process.
4. The loan overall process accurate wind control and management system based on massive big data and core algorithm according to claim 1, characterized in that: the behavior identification module (301), the credit rating module (302), the risk assessment module (303) and the post-credit warning module (304) are sequentially connected through Ethernet communication and run in parallel; the behavior recognition module (301) is used for monitoring the behaviors of users and institutions in the loan business process and recognizing the behavior with risks; the credit rating module (302) is used for rating the credit condition of users and organizations by combining big data analysis and prediction of a training model; the risk evaluation module (303) is used for carrying out omnibearing analysis and judgment on risks and risk conditions existing in the whole loan business process; and the post-loan early warning module (304) is used for carrying out early warning analysis on the risk condition existing in the post-processing process of the loan service according to the risk assessment condition of multiple parties.
5. The loan overall process accurate wind control and management system based on massive big data and core algorithm according to claim 4, characterized in that: the risk assessment module (303) comprises a risk simulation module (3031), a measurement early warning module (3032), a dynamic monitoring module (3033) and a process optimization module (3034); a signal output end of the risk simulation module (3031) is connected with a signal input end of the measurement early warning module (3032), a signal output end of the measurement early warning module (3032) is connected with a signal input end of the dynamic monitoring module (3033), and a signal output end of the dynamic monitoring module (3033) is connected with a signal input end of the process optimization module (3034); the risk simulation module (3031) is used for performing simulation analysis on the existing risk through an algorithm model; the measurement early warning module (3032) is used for measuring the risk and the risk size obtained through simulation analysis and sending early warning to related personnel when the risk size reaches a certain degree; the dynamic monitoring module (3033) is used for dynamically monitoring risks in the whole process of the loan business on the basis of big data analysis; and the flow optimization module (3034) is used for optimizing and adjusting the business flow with risk according to the risk condition of the comprehensive analysis.
6. The loan overall process accurate wind control and management system based on massive big data and core algorithm according to claim 1, characterized in that: the signal output end of the index calculation module (401) is connected with the signal input end of the operation checking module (402), the signal output end of the operation checking module (402) is connected with the signal input end of the wind control decision module (403), and the signal output end of the wind control decision module (403) is connected with the signal input end of the comprehensive report module (404); the index calculation module (401) is used for carrying out real-time flow calculation and graph calculation on the loan service related data according to a preset index center; the operation auditing module (402) is used for auditing and auditing the loan business case on an operation auditing platform in a random extraction mode; the wind control decision module (403) is used for integrating the result analysis of rule decision and model decision to select and implement risk decision; the comprehensive report module (404) is used for collating the full-process data of data analysis, risk prediction, operation examination and wind control decision related to the loan service and forming a corresponding report.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110649992.0A CN113393316B (en) | 2021-06-10 | 2021-06-10 | Loan overall process accurate wind control and management system based on massive big data and core algorithm |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110649992.0A CN113393316B (en) | 2021-06-10 | 2021-06-10 | Loan overall process accurate wind control and management system based on massive big data and core algorithm |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN113393316A CN113393316A (en) | 2021-09-14 |
| CN113393316B true CN113393316B (en) | 2023-02-03 |
Family
ID=77620356
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202110649992.0A Active CN113393316B (en) | 2021-06-10 | 2021-06-10 | Loan overall process accurate wind control and management system based on massive big data and core algorithm |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN113393316B (en) |
Families Citing this family (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116862643A (en) * | 2023-06-25 | 2023-10-10 | 福建润楼数字科技有限公司 | Automatic wind control feature screening method for multi-channel fund integration credit business |
| CN117172908A (en) * | 2023-09-05 | 2023-12-05 | 中铁商业保理有限公司 | Business loan stage decision-making auxiliary system based on consumption bill analysis |
Citations (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109784466A (en) * | 2019-01-25 | 2019-05-21 | 潍坊医学院 | A kind of adaptive dynamic learning particle swarm optimization algorithm |
Family Cites Families (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107481135A (en) * | 2017-08-16 | 2017-12-15 | 广东工业大学 | A kind of personal credit evaluation method and system based on BP neural network |
| CN110458687A (en) * | 2019-07-05 | 2019-11-15 | 平安银行股份有限公司 | The automatic measures and procedures for the examination and approval of decision, device and computer readable storage medium |
| CN112348656A (en) * | 2020-09-29 | 2021-02-09 | 百维金科(上海)信息科技有限公司 | BA-WNN-based personal loan credit scoring method |
| CN112661259A (en) * | 2020-12-15 | 2021-04-16 | 海天水务集团股份公司 | Self-adaptive control method and system for aeration process |
-
2021
- 2021-06-10 CN CN202110649992.0A patent/CN113393316B/en active Active
Patent Citations (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109784466A (en) * | 2019-01-25 | 2019-05-21 | 潍坊医学院 | A kind of adaptive dynamic learning particle swarm optimization algorithm |
Also Published As
| Publication number | Publication date |
|---|---|
| CN113393316A (en) | 2021-09-14 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109034194B (en) | Transaction fraud behavior deep detection method based on feature differentiation | |
| CN116342259A (en) | Automatic user credit rating method and device, electronic equipment and medium | |
| Wang et al. | Multi-classification assessment of bank personal credit risk based on multi-source information fusion | |
| CN113393316B (en) | Loan overall process accurate wind control and management system based on massive big data and core algorithm | |
| CN111178675A (en) | LR-Bagging algorithm-based electric charge recycling risk prediction method, system, storage medium and computer equipment | |
| CN111127178A (en) | Data processing method and device, storage medium and electronic equipment | |
| CN113139687B (en) | Method and device for predicting credit card user default | |
| CN109829721B (en) | Online transaction multi-subject behavior modeling method based on heterogeneous network characterization learning | |
| CN110895758A (en) | Screening method, device and system for credit card account with cheating transaction | |
| CN112700324A (en) | User loan default prediction method based on combination of Catboost and restricted Boltzmann machine | |
| CN109102396A (en) | A kind of user credit ranking method, computer equipment and readable medium | |
| CN113298121A (en) | Message sending method and device based on multi-data source modeling and electronic equipment | |
| CN111861487A (en) | Financial transaction data processing method, and fraud monitoring method and device | |
| CN115908022A (en) | Abnormal transaction risk early warning method and system based on network modeling | |
| CN112950359B (en) | User identification method and device | |
| CN112634048A (en) | Anti-money laundering model training method and device | |
| CN117575595A (en) | Payment risk identification method, device, computer equipment and storage medium | |
| CN111475566A (en) | Identification method and device for block chain fund suspicious transaction mode | |
| KR102336462B1 (en) | Apparatus and method of credit rating | |
| CN114612239A (en) | Stock public opinion monitoring and wind control system based on algorithm, big data and artificial intelligence | |
| Li | Credit card fraud identification based on unbalanced data set based on fusion model | |
| CN111932368A (en) | Credit card issuing system and construction method and device thereof | |
| CN112823502A (en) | Real-time feedback service configured for resource access rules | |
| CN117172910A (en) | Credit evaluation method and device based on EBM model, electronic equipment and storage medium | |
| CN118333737A (en) | Method for constructing retail credit risk prediction model and consumer credit business Scorebetai model |
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 | ||
| CB03 | Change of inventor or designer information |
Inventor after: Luo Siyang Inventor after: Luo Zhongming Inventor before: Luo Zhongming |
|
| CB03 | Change of inventor or designer information | ||
| TA01 | Transfer of patent application right |
Effective date of registration: 20230111 Address after: 201100 Room 302, No. 3, Lane 59, Lane 3, Fuqiang street, Qibao town, Minhang District, Shanghai Applicant after: Luo Siyang Address before: 730000 Zhonghai Heshan County, liujiabao, Lanzhou City, Gansu Province Applicant before: Luo Zhongming |
|
| TA01 | Transfer of patent application right | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |