CN111951104A - Risk conduction early warning method based on associated graph - Google Patents
Risk conduction early warning method based on associated graph Download PDFInfo
- Publication number
- CN111951104A CN111951104A CN202010857277.1A CN202010857277A CN111951104A CN 111951104 A CN111951104 A CN 111951104A CN 202010857277 A CN202010857277 A CN 202010857277A CN 111951104 A CN111951104 A CN 111951104A
- Authority
- CN
- China
- Prior art keywords
- clients
- client
- community
- blacklist
- risk
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 238000000034 method Methods 0.000 title claims abstract description 42
- 238000004364 calculation method Methods 0.000 claims description 15
- 238000003062 neural network model Methods 0.000 claims description 13
- 238000010801 machine learning Methods 0.000 claims description 9
- 238000011835 investigation Methods 0.000 claims description 3
- 230000002596 correlated effect Effects 0.000 claims description 2
- 230000026683 transduction Effects 0.000 claims 8
- 238000010361 transduction Methods 0.000 claims 8
- 238000004458 analytical method Methods 0.000 description 2
- 230000006399 behavior Effects 0.000 description 2
- 230000006866 deterioration Effects 0.000 description 2
- 230000004075 alteration Effects 0.000 description 1
- 238000011156 evaluation Methods 0.000 description 1
- 238000012502 risk assessment Methods 0.000 description 1
- 238000006467 substitution reaction Methods 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
- 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
- G06Q30/00—Commerce
- G06Q30/02—Marketing; Price estimation or determination; Fundraising
- G06Q30/0201—Market modelling; Market analysis; Collecting market data
Landscapes
- Business, Economics & Management (AREA)
- Accounting & Taxation (AREA)
- Finance (AREA)
- Engineering & Computer Science (AREA)
- Strategic Management (AREA)
- Development Economics (AREA)
- Theoretical Computer Science (AREA)
- Marketing (AREA)
- Physics & Mathematics (AREA)
- General Business, Economics & Management (AREA)
- General Physics & Mathematics (AREA)
- Economics (AREA)
- Entrepreneurship & Innovation (AREA)
- Technology Law (AREA)
- Data Mining & Analysis (AREA)
- Game Theory and Decision Science (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention relates to a risk conduction early warning method based on an associated map, which comprises the following steps: acquiring all incidence relations of the clients and establishing incidence maps of the clients; classifying the clients to obtain white list clients and black list clients, and early warning the black list clients; predicting risk customers in the associated map according to the associated map and blacklist customers obtained by classification, and early warning the risk customers; and carrying out community division on the clients by using a louvain algorithm, and carrying out early warning on the communities according to the community blacklist client coverage rate and the community density. According to the method and the system, early warning of a single client, early warning of associated risk clients in the association map and early warning of community risks are achieved by constructing the association map of the client, classifying the client, predicting the risk of the associated client, evaluating the risk of the community and the like. The problem that the relationship between the clients cannot be predicted, and the risk of the community cannot be evaluated and early-warned in the prior art is solved.
Description
Technical Field
The invention relates to the technical field of risk conduction, in particular to a risk conduction early warning method based on an associated map.
Background
Under the traditional business mode of the bank, the customer insight usually takes a single customer as a research object and pays less attention to the customer relationship network. In the aspect of customer relationship, management is mainly performed on customers with direct relationship, such as group customers, guarantors and the like, and research on a customer relationship network is lacked.
Because the customer insight usually uses a single customer as a research object, the existing analysis method for risk conduction between customers by a bank can only realize risk assessment on customers with one-to-one direct relationship, for example, when a customer a has a risk (for example, a default behavior), only can assess whether a customer B having a direct relationship with the customer a has a risk and a risk probability, but cannot predict the relationship of other customers having an indirect relationship with the customer a, and even cannot assess and early-warn risk conduction results (generally including risk conduction probability) of other customers caused by the risk of the customer a; the risk influence on the community where the client a is located cannot be evaluated and warned.
Therefore, it is necessary to provide a risk conduction early warning method based on an association graph to solve the problem that the relationship between clients cannot be predicted and the risk of a community cannot be evaluated and early warned in the prior art.
Disclosure of Invention
The invention aims to provide a risk conduction early warning method based on an association graph, which aims to solve the problems that the relationship prediction between clients and the evaluation and early warning of the risk of a community cannot be carried out in the prior art.
In order to solve the problems in the prior art, the invention provides a risk conduction early warning method based on an associated map, which comprises the following steps:
acquiring all incidence relations of the clients and establishing incidence maps of the clients;
classifying the clients to obtain white list clients and black list clients, and early warning the black list clients;
predicting risk customers in the associated map according to the associated map and blacklist customers obtained by classification, and early warning the risk customers;
and carrying out community division on the clients by using a louvain algorithm, and carrying out early warning on the communities according to the community blacklist client coverage rate and the community density.
Optionally, in the risk conduction early warning method based on the association map, the data source of the association relationship includes: bank data, business data, credit data, and credit investigation data.
Optionally, in the risk conduction early warning method based on the association map, a graph neural network model or a machine learning model is used for classifying the customers.
Optionally, in the risk conduction early warning method based on the association map, a manner of classifying the customers is as follows:
giving out the probability of each client being a blacklist by adopting a graph neural network model or a machine learning model;
and judging the probability of each client as a blacklist by the graph neural network model or the machine learning model, if the probability of each client as the blacklist is greater than 85%, determining the client as the blacklist, and otherwise, determining the client as the white list.
Optionally, in the risk conduction early warning method based on the association map, predicting risk customers in the association map includes the following steps:
judging whether the client is directly associated with the blacklist client obtained by classification, if so, judging the client to be a risk client, and performing early warning; and if the indirect correlation exists, calculating the probability of edges existing between the indirectly correlated clients and the blacklist clients obtained by classification by adopting a relation prediction calculation formula, and if the probability of the edges existing is greater than a probability threshold value, judging the clients to be risk clients and carrying out early warning.
Optionally, in the risk conduction early warning method based on the association map, the relationship prediction calculation formula is as follows:
wherein, a (x, y) is a probability that an edge exists between any clients, n (u) is a set of nodes adjacent to a node u, and the node u is a node formed by any client in the association graph.
Optionally, in the risk conduction early warning method based on the association graph, the early warning of the community further includes the following steps:
presetting a community blacklist client coverage rate threshold and a community density threshold;
calculating the client coverage rate and the community density of each community blacklist;
and when the calculated community blacklist customer coverage rate is larger than a community blacklist customer coverage rate threshold value and/or the calculated community density is larger than a community density threshold value, early warning is carried out.
Optionally, in the risk conduction early warning method based on the association map,
the calculation formula of the community blacklist client coverage rate is as follows: pc is the community blacklist client coverage rate, B is the quantity of blacklist clients in the community, and A is the quantity of all clients in the community;
the calculation formula of the community density is as follows: and D is S/A, wherein D is the community density, S is the number of associated edges in the community map, and A is the number of all clients in the community.
Optionally, in the risk conduction early warning method based on the association map,
all the association relations of the client comprise all the association relations of the public client and all the association relations of the private client.
Optionally, in the risk conduction early warning method based on the association map, the risk conduction early warning method based on the association map further includes calculating a comprehensive risk probability of the customer, where the calculation method is as follows:
p is Pc + Pn, wherein P is the comprehensive risk probability, Pc is the coverage rate of the blacklist client in the community where the client is located, and Pn is the probability of being the blacklist when the client is classified;
and when the probability of being the blacklist is more than 85% when the clients are classified, the clients are determined to be blacklist clients, early warning is carried out on the blacklist clients, and meanwhile the probability and the comprehensive danger probability of the blacklist clients being the blacklist are pushed.
According to the risk conduction early warning method based on the association graph, the association graph of the client is constructed by utilizing the association relation data of the client, and early warning of a single client, early warning of associated risk clients in the association graph and early warning of community risks are achieved by classifying the client, predicting the risks of the associated client, evaluating the risks of communities and the like. The problem that the relationship between the clients cannot be predicted, and the risk of the community cannot be evaluated and early-warned in the prior art is solved.
Drawings
Fig. 1 is a flowchart of a risk conduction early warning method according to an embodiment of the present invention.
Detailed Description
The following describes in more detail embodiments of the present invention with reference to the schematic drawings. The advantages and features of the present invention will become more apparent from the following description. It is to be noted that the drawings are in a very simplified form and are not to precise scale, which is merely for the purpose of facilitating and distinctly claiming the embodiments of the present invention.
Hereinafter, if the method described herein comprises a series of steps, the order of such steps presented herein is not necessarily the only order in which such steps may be performed, and some of the described steps may be omitted and/or some other steps not described herein may be added to the method.
Under a traditional banking business mode, a client insights usually take a single client as a research object, so that the existing analysis method for risk conduction between clients by a bank can only evaluate the risk of the client with one-to-one direct relationship, for example, when a client a has a risk (for example, a default behavior), only a client B having a direct relationship with the client a can be evaluated whether the client B has the risk and the risk probability, but cannot predict the relationship of other clients having an indirect relationship with the client a, and even cannot evaluate and warn the risk conduction result (generally including the risk conduction probability) of other clients caused by the risk of the client a; the risk influence on the community where the client a is located cannot be evaluated and warned.
Therefore, it is necessary to provide a risk conduction early warning method based on an association graph, as shown in fig. 1, fig. 1 is a flowchart of the risk conduction early warning method provided in the embodiment of the present invention, where the risk conduction early warning method includes the following steps:
acquiring all incidence relations of the clients and establishing incidence maps of the clients;
classifying the clients to obtain white list clients and black list clients, and early warning the black list clients;
predicting risk customers in the associated map according to the associated map and blacklist customers obtained by classification, and early warning the risk customers;
and carrying out community division on the clients by using a louvain algorithm, and carrying out early warning on the communities according to the community blacklist client coverage rate and the community density.
According to the method and the system, the association map of the client is constructed by utilizing the association relation data of the client, and then early warning of a single client, early warning of associated risk clients in the association map and early warning of community risks are realized by classifying the client, predicting the risks of the associated client, evaluating the risks of the community and the like. The problem that the relationship between the clients cannot be predicted, and the risk of the community cannot be evaluated and early-warned in the prior art is solved.
Further, the data source of the association relationship includes: bank data, business data, credit data, and credit investigation data. All the association relations of the client comprise all the association relations of a public client and all the association relations of a private client, for example, natural person relations, enterprise legal person relations, relations among enterprises and the like are included.
Furthermore, the invention uses a graph neural network model or a machine learning model to classify the clients, and preferably uses the graph neural network model to classify the clients.
Specifically, the manner of classifying the customers is as follows: giving out the probability of each client being a blacklist by adopting a graph neural network model or a machine learning model; and judging the probability of each client as a blacklist by the graph neural network model or the machine learning model, if the probability of each client as the blacklist is greater than 85%, determining the client as the blacklist, and otherwise, determining the client as the white list.
In one embodiment, the probability that each client is a blacklist is given by using a graph neural network model, and the specific operation mode is as follows: firstly, basic attribute characteristics [ x1, x2, …, xn ] of each client are included, then characteristic engineering is carried out to obtain characteristics [ y1, y2, …, yn ], then correlation map characteristics are considered, finally, the characteristic engineering and the correlation map characteristics are recorded into the graph neural network model, and the probability that each client is a blacklist is given by the graph neural network model;
the following table is an example of the relevant map features:
preferably, the method for predicting risk customers in the association graph comprises the following steps:
judging whether the client is directly associated with the blacklist client obtained by classification, if so, judging the client to be a risk client, and performing early warning; if the indirect association exists, calculating the probability of the edge existing between the indirectly associated client and the blacklist client obtained by classification by using a relational prediction calculation formula, if the probability of the edge existing is greater than a probability threshold value, judging the indirectly associated client to be a risk client, and performing early warning, wherein the probability threshold value can be 70%, namely, when the probability of the edge existing between the indirectly associated client and the blacklist client obtained by classification is greater than 70%, early warning is performed on the indirectly associated client.
Optionally, in the risk conduction early warning method based on the association map, the relationship prediction calculation formula is as follows:
wherein, a (x, y) is a probability that an edge exists between any clients, n (u) is a node set adjacent to a node u, the node u is a node formed by any client in the association graph, and the node u is a node connecting the node x and the node y.
The prediction mode adopted in the invention can predict the relationship between the hidden existence edge and the indirect existence edge, and can also predict the possible existence edge relationship in the future.
Preferably, in the risk conduction early warning method based on the association map, a louvain algorithm is adopted to perform community division on the clients, and the formula of the community division is as follows:
wherein L is the sum of all edges in the association map; l iscIs the sum of all edges in the community; kcIs the sum of all the degrees of nodes in the community. According to the invention, the customer is divided into communities by adopting a louvain algorithm, and the community map of each community is constructed, so that each community comprises a plurality of customers, and no customer intersection exists between each community.
Further, the early warning of the community further comprises the following steps: presetting a community blacklist client coverage rate threshold and a community density threshold; calculating the client coverage rate and the community density of each community blacklist; and when the calculated community blacklist customer coverage rate is larger than a community blacklist customer coverage rate threshold value and/or the calculated community density is larger than a community density threshold value, early warning is carried out. The calculation formula of the community blacklist client coverage rate is as follows: pc is the community blacklist client coverage rate, B is the quantity of blacklist clients in the community, and A is the quantity of all clients in the community; the calculation formula of the community density is as follows: and D is S/A, wherein D is the community density, S is the number of associated edges in the community map, and A is the number of all clients in the community. In one embodiment, the community blacklist client coverage threshold may be, for example, 20%, and the community density threshold is, for example, 2 times of the entire association map density, that is, when the calculated community blacklist client coverage is greater than 20%, and/or the calculated community density is greater than 2 times of the entire association map density, the warning is performed.
Generally, the invention can also make the community in a community shape, if the clients in the community are on a circle link, the community can be made in a circle shape, and similar shapes can also be a triangle shape, a pyramid shape and the like.
Further, for the community number, the client number with the largest PageRank value is used as the community number, and the PageRank is the node with the largest importance and relevance in the community map. Generally, the change of the community type is used as the standard for whether the community map changes. PageRank is calculated as follows:
where d is a coefficient, Tn represents a node, and C (Tn) represents the degree of departure of the node Tn.
Preferably, for a new customer, the risk conduction early warning method based on the association graph further includes calculating a comprehensive risk probability of the customer, and the calculation method includes:
p is Pc + Pn, wherein P is the comprehensive risk probability, Pc is the coverage rate of the blacklist client in the community where the client is located, and Pn is the probability of being the blacklist when the client is classified;
and when the probability of being the blacklist is more than 85% when the clients are classified, the clients are determined to be blacklist clients, early warning is carried out on the blacklist clients, and meanwhile the probability and the comprehensive danger probability of the blacklist clients being the blacklist are pushed.
According to the method, for old customers, the information such as bad accounts can be known according to external data in the association map, the information is converted from a white list to a black list, the black list customers directly warn, then directly associated customers are checked, the loss is stopped in time, then indirectly associated customers are checked, the probability of possible deterioration is given by using a relation prediction calculation formula, and the indirectly associated customers with the deterioration probability of more than 70% are warned. And for the new client, predicting the probability of the new client being the blacklist in time, judging whether the new client is credited or not, and the like. Therefore, the new and old customers can be judged in time, and the air control gate-closing capability is effectively improved.
In summary, in the risk conduction early warning method based on the association graph provided by the invention, the association graph of the client is constructed by using the association relation data of the client, and then the early warning of a single client, the early warning of the associated risk client in the association graph and the early warning of the community risk are realized by classifying the client, performing risk prediction on the associated client, evaluating the risk of the community and the like. The problem that the relationship between the clients cannot be predicted, and the risk of the community cannot be evaluated and early-warned in the prior art is solved.
The above description is only a preferred embodiment of the present invention, and does not limit the present invention in any way. It will be understood by those skilled in the art that various changes, substitutions and alterations can be made herein without departing from the spirit and scope of the invention as defined by the appended claims.
Claims (10)
1. A risk conduction early warning method based on an association map is characterized by comprising the following steps:
acquiring all incidence relations of the clients and establishing incidence maps of the clients;
classifying the clients to obtain white list clients and black list clients, and early warning the black list clients;
predicting risk customers in the associated map according to the associated map and blacklist customers obtained by classification, and early warning the risk customers;
and carrying out community division on the clients by using a louvain algorithm, and carrying out early warning on the communities according to the community blacklist client coverage rate and the community density.
2. The risk transduction early warning method based on the associative graph according to claim 1, wherein the data source of the associative relationship comprises: bank data, business data, credit data, and credit investigation data.
3. The relevance graph-based risk-conducting warning method according to claim 1, wherein a graph neural network model or a machine learning model is used to classify the clients.
4. The risk transduction pre-warning method based on the associative graph as set forth in claim 3, wherein the clients are classified in a manner of:
giving out the probability of each client being a blacklist by adopting a graph neural network model or a machine learning model;
and judging the probability of each client as a blacklist by the graph neural network model or the machine learning model, if the probability of each client as the blacklist is greater than 85%, determining the client as the blacklist, and otherwise, determining the client as the white list.
5. The relevance graph-based risk transduction early warning method according to claim 1, wherein predicting risk customers in the relevance graph comprises the steps of:
judging whether the client is directly associated with the blacklist client obtained by classification, if so, judging the client to be a risk client, and performing early warning; and if the indirect correlation exists, calculating the probability of edges existing between the indirectly correlated clients and the blacklist clients obtained by classification by adopting a relation prediction calculation formula, and if the probability of the edges existing is greater than a probability threshold value, judging the clients to be risk clients and carrying out early warning.
6. The risk transduction early warning method based on the associative graph according to claim 5, wherein the relational prediction calculation formula is:
wherein, a (x, y) is a probability that an edge exists between any clients, n (u) is a set of nodes adjacent to a node u, and the node u is a node formed by any client in the association graph.
7. The risk transduction early warning method based on the association graph as claimed in claim 1, wherein the early warning of the community further comprises the steps of:
presetting a community blacklist client coverage rate threshold and a community density threshold;
calculating the client coverage rate and the community density of each community blacklist;
and when the calculated community blacklist customer coverage rate is larger than a community blacklist customer coverage rate threshold value and/or the calculated community density is larger than a community density threshold value, early warning is carried out.
8. The risk transduction pre-warning method based on correlation graph according to claim 7,
the calculation formula of the community blacklist client coverage rate is as follows: pc is the community blacklist client coverage rate, B is the quantity of blacklist clients in the community, and A is the quantity of all clients in the community;
the calculation formula of the community density is as follows: and D is S/A, wherein D is the community density, S is the number of associated edges in the community map, and A is the number of all clients in the community.
9. The risk transduction pre-warning method based on correlation graph according to claim 1,
all the association relations of the client comprise all the association relations of the public client and all the association relations of the private client.
10. The association graph-based risk transduction early warning method according to claim 1, further comprising calculating a comprehensive risk probability of the customer by:
p is Pc + Pn, wherein P is the comprehensive risk probability, Pc is the coverage rate of the blacklist client in the community where the client is located, and Pn is the probability of being the blacklist when the client is classified;
and when the probability of being the blacklist is more than 85% when the clients are classified, the clients are determined to be blacklist clients, early warning is carried out on the blacklist clients, and meanwhile the probability and the comprehensive danger probability of the blacklist clients being the blacklist are pushed.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010857277.1A CN111951104A (en) | 2020-08-24 | 2020-08-24 | Risk conduction early warning method based on associated graph |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010857277.1A CN111951104A (en) | 2020-08-24 | 2020-08-24 | Risk conduction early warning method based on associated graph |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN111951104A true CN111951104A (en) | 2020-11-17 |
Family
ID=73359597
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202010857277.1A Pending CN111951104A (en) | 2020-08-24 | 2020-08-24 | Risk conduction early warning method based on associated graph |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN111951104A (en) |
Cited By (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112598496A (en) * | 2020-12-15 | 2021-04-02 | 深圳前海微众银行股份有限公司 | Wind control blacklist setting method and device, terminal equipment and readable storage medium |
| CN112699249A (en) * | 2020-12-31 | 2021-04-23 | 上海浦东发展银行股份有限公司 | Knowledge graph-based information processing method, device, equipment and storage medium |
| CN112785423A (en) * | 2021-02-07 | 2021-05-11 | 撼地数智(重庆)科技有限公司 | Method, device, equipment and storage medium for mining fraud risk node |
| CN113516553A (en) * | 2021-07-28 | 2021-10-19 | 中国建设银行股份有限公司 | Credit risk early warning method and device |
| CN113989005A (en) * | 2021-10-27 | 2022-01-28 | 航天信息股份有限公司 | Tax risk enterprise mining method and device |
Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109657918A (en) * | 2018-11-19 | 2019-04-19 | 平安科技（深圳）有限公司 | Method for prewarning risk, device and the computer equipment of association assessment object |
| CN110362639A (en) * | 2019-06-27 | 2019-10-22 | 上海淇馥信息技术有限公司 | A kind of method for prewarning risk, device, electronic equipment calculating analysis based on demographic associations |
| CN110555564A (en) * | 2019-09-06 | 2019-12-10 | 中国农业银行股份有限公司 | Method and device for predicting client associated risk |
| CN110647590A (en) * | 2019-09-23 | 2020-01-03 | 税友软件集团股份有限公司 | Target community data identification method and related device |
| CN110717824A (en) * | 2019-10-17 | 2020-01-21 | 北京明略软件系统有限公司 | Method and device for conducting and calculating risk of public and guest groups by bank based on knowledge graph |
| CN111309824A (en) * | 2020-02-18 | 2020-06-19 | 中国工商银行股份有限公司 | Entity relationship map display method and system |
| CN111340611A (en) * | 2020-02-20 | 2020-06-26 | 中国建设银行股份有限公司 | Risk early warning method and device |
-
2020
- 2020-08-24 CN CN202010857277.1A patent/CN111951104A/en active Pending
Patent Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109657918A (en) * | 2018-11-19 | 2019-04-19 | 平安科技（深圳）有限公司 | Method for prewarning risk, device and the computer equipment of association assessment object |
| CN110362639A (en) * | 2019-06-27 | 2019-10-22 | 上海淇馥信息技术有限公司 | A kind of method for prewarning risk, device, electronic equipment calculating analysis based on demographic associations |
| CN110555564A (en) * | 2019-09-06 | 2019-12-10 | 中国农业银行股份有限公司 | Method and device for predicting client associated risk |
| CN110647590A (en) * | 2019-09-23 | 2020-01-03 | 税友软件集团股份有限公司 | Target community data identification method and related device |
| CN110717824A (en) * | 2019-10-17 | 2020-01-21 | 北京明略软件系统有限公司 | Method and device for conducting and calculating risk of public and guest groups by bank based on knowledge graph |
| CN111309824A (en) * | 2020-02-18 | 2020-06-19 | 中国工商银行股份有限公司 | Entity relationship map display method and system |
| CN111340611A (en) * | 2020-02-20 | 2020-06-26 | 中国建设银行股份有限公司 | Risk early warning method and device |
Non-Patent Citations (1)
| Title |
|---|
| 朱艳敏; 陈超: ""Z-score 模型最优分割点的确定方法比较"", 《南方金融》, pages 74 - 77 * |
Cited By (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112598496A (en) * | 2020-12-15 | 2021-04-02 | 深圳前海微众银行股份有限公司 | Wind control blacklist setting method and device, terminal equipment and readable storage medium |
| CN112598496B (en) * | 2020-12-15 | 2024-04-30 | 深圳前海微众银行股份有限公司 | Wind control blacklist setting method and device, terminal equipment and readable storage medium |
| CN112699249A (en) * | 2020-12-31 | 2021-04-23 | 上海浦东发展银行股份有限公司 | Knowledge graph-based information processing method, device, equipment and storage medium |
| CN112785423A (en) * | 2021-02-07 | 2021-05-11 | 撼地数智(重庆)科技有限公司 | Method, device, equipment and storage medium for mining fraud risk node |
| CN113516553A (en) * | 2021-07-28 | 2021-10-19 | 中国建设银行股份有限公司 | Credit risk early warning method and device |
| CN113989005A (en) * | 2021-10-27 | 2022-01-28 | 航天信息股份有限公司 | Tax risk enterprise mining method and device |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Ali et al. | A data-driven approach for multi-scale GIS-based building energy modeling for analysis, planning and support decision making | |
| CN111951104A (en) | Risk conduction early warning method based on associated graph | |
| Goh et al. | Incorporating the rough sets theory into travel demand analysis | |
| CN110188198B (en) | Anti-fraud method and device based on knowledge graph | |
| AU2018101946A4 (en) | Geographical multivariate flow data spatio-temporal autocorrelation analysis method based on cellular automaton | |
| Feng et al. | Using exploratory regression to identify optimal driving factors for cellular automaton modeling of land use change | |
| Xue et al. | Spatial analysis with preference specification of latent decision makers for criminal event prediction | |
| CN115545450A (en) | Carbon emission collaborative prediction method based on digital twinning | |
| TWI464700B (en) | Method and device for credit default prediction | |
| CN112700324A (en) | User loan default prediction method based on combination of Catboost and restricted Boltzmann machine | |
| Xu et al. | Novel key indicators selection method of financial fraud prediction model based on machine learning hybrid mode | |
| AU2006239438A1 (en) | Test mining systems and methods for early detection and warning | |
| Shil et al. | Forecasting electric vehicle adoption in the USA using machine learning models | |
| US7979366B2 (en) | System and method for coarse-classing variables in a propensity model | |
| CN118133274A (en) | Information security management and monitoring method and system based on big data | |
| KR102499182B1 (en) | Loan regular auditing system using artificia intellicence | |
| TWM622331U (en) | System and device for risk prediction therefor | |
| Plewis et al. | Using response propensity models to improve the quality of response data in longitudinal studies | |
| KR102499183B1 (en) | Loan regular auditing system using artificia intellicence | |
| Lin et al. | Mitigating subgroup unfairness in machine learning classifiers: A data-driven approach | |
| Yunsong et al. | An XGboost Algorithm Based Model for Financial Risk Prediction. | |
| Xu et al. | An XGboost Algorithm Based Model for Financial Risk Prediction | |
| Pedra et al. | Windstorm Economic Impacts on the Spanish Resilience: A Machine Learning Real-Data Approach | |
| Riedel | The problems of assessing transnational mobility: Identifying latent groups of immigrants in Germany using factor mixture analysis | |
| KR102478555B1 (en) | Server selecting candidates for the crackdown site for agricultural products and operating method thereof |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| RJ01 | Rejection of invention patent application after publication | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20201117 |