A supply chain financial risk assessment method and system

By establishing a transaction matrix and utilizing neural network models, the problem of incomplete construction of supply chain financial risk assessment parameters in the existing technology is solved, efficient and accurate risk assessment is achieved, and the security and efficiency of supply chain finance are improved.

CN118628114BActive Publication Date: 2024-11-12SICHUAN SILUYIGOU TECH CO LTD
View PDF 14 Cites 0 Cited by

Patent Information

Application Number
CN202411102744.4
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2024-08-13
Publication Date
2024-11-12
Estimated Expiration
2044-08-13

AI Technical Summary

Technical Problem

In the supply chain financial risk assessment, it is difficult to effectively construct sufficiently representative and reliable parameters in the supply chain financial risk assessment, resulting in insufficient comprehensive and accurate risk assessment.

Method used

By obtaining supply chain transaction data during the time period, three types of transaction matrices are established: the first matrix represents the transaction relationship between nodes, the second matrix represents the association relationship, and the third matrix represents the risk degree, and the trained neural network model is used to judge Supply chain financial risk level.

Benefits of technology

It has achieved efficient and automated financial risk assessment, which can dynamically capture and analyze supply chain changes, timely identify potential risks, improve the reliability and accuracy of risk assessment, and promote the security and efficiency of supply chain finance.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118628114B_ABST
    Figure CN118628114B_ABST
Patent Text Reader

Abstract

The present invention provides a supply chain financial risk assessment method and system, which relates to the technical field of supply chain risk assessment, and aims to achieve more reliable supply chain financial risk assessment through more reasonable parameter extraction and construction methods, including obtaining transaction data on the supply chain within a time period; establishing a first transaction matrix; establishing a second transaction matrix; establishing a third transaction matrix; and judging the supply chain financial risk level through a trained recognition model. The present invention has the advantages of high cost performance of data extraction, comprehensive feature expression and accurate risk assessment results.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to the technical field of supply chain risk assessment, and in particular to a supply chain financial risk assessment method and system. Background Art

[0002] Supply chain finance risk assessment is a key link in supply chain finance, which helps to effectively monitor and prevent risks for various supply chain companies.

[0003] In supply chain finance risk assessment, supply chain finance risk assessment involves multiple factors and the risk-related data is relatively complex. Therefore, in the existing technology, whether it is through neural networks or real-time threshold monitoring, it is challenging to effectively construct sufficiently representative and reliable parameters to achieve a more comprehensive risk assessment.

[0004] Therefore, it is urgent to design a supply chain finance risk assessment method to achieve a more reliable supply chain finance risk assessment through more reasonable parameter extraction and its construction method. Summary of the invention

[0005] The purpose of the present invention is to provide a supply chain finance risk assessment method and system, which can achieve more reliable supply chain finance risk assessment through more reasonable parameter extraction and its construction method.

[0006] The embodiments of the present invention are implemented by the following technical solutions:

[0007] A supply chain finance risk assessment method comprises the following steps:

[0008] Obtain transaction data on the supply chain within a time period;

[0009] Establishing the first trading matrix , the first transaction matrix Line The column elements represent the first Node and The transaction relationship between nodes. The larger the value, the more transactions.

[0010] Establishing the Second Trading Matrix , the second transaction matrix Line The column elements represent the first Node and The smaller the value, the closer the relationship between nodes.

[0011] Establishing the third trading matrix , the third transaction matrix is ​​a row matrix, whose The element represents the first The risk level of the node, and the smaller the value, the lower the risk;

[0012] The supply chain finance risk level is determined based on the first transaction matrix, the second transaction matrix and the third transaction matrix through a trained recognition model.

[0013] Preferably, the method for establishing the first transaction matrix is:

[0014]

[0015] = ;

[0016] in, is the first Line The elements in the column represent the first Node and The number of transactions between nodes, is the first Line The elements in the column represent the first Node and The transaction amount coefficient between nodes, , , and are weights, is the total number of nodes, the matrix and matrix The diagonal elements of are all 0.

[0017] Preferably, the supply chain Node and The method for obtaining the transaction amount coefficient between nodes is:

[0018] ;

[0019] in, For the supply chain Node and The total amount of transactions between nodes, For the supply chain Node and The total amount of transactions between nodes.

[0020] Preferably, the method for establishing the second transaction matrix is:

[0021] ;

[0022] in, The second transaction matrix Middle Line The elements in the column represent the first Node and The relationship between nodes, , , is the total number of nodes;

[0023] ;

[0024] in, For the supply chain Node and Historical transaction frequency between nodes, For the supply chain Node and The number of historical transaction product types between nodes, For the supply chain Node and The historical transaction time interval difference parameter between nodes, For the supply chain Node and The average amount of each historical transaction between nodes, , and are weights respectively, and the diagonal elements of the second transaction matrix are all 0.

[0025] Preferably, the supply chain Node and The method for obtaining the historical transaction time interval parameters between nodes is:

[0026] Based on historical transactions, the first Node and Continuity between nodes transactions as calculation samples;

[0027] Get :

[0028] ;

[0029] in, Represents the first Node and The number of nodes between Transactions and The time interval between transactions, Represents the first Node and The number of nodes between Transactions and The time interval between transactions.

[0030] Preferably, the third transaction matrix is ​​established The method is:

[0031] The establishment of the third transaction matrix for:

[0032] ;

[0033] in, The establishment of the third transaction matrix No. elements, representing the first The node's historical transaction parameters, , is the total number of nodes;

[0034] ;

[0035] in, and They are respectively Node close The payment on-time rate and average payment delay time of each transaction, and They are respectively Node close The on-time delivery rate and average delivery delay time of each transaction, For the supply chain Node close The number of violations and defaults in transactions, , and are weights, .

[0036] Preferably, the recognition model includes an input layer, a feature extraction layer, a fully connected layer and an output layer;

[0037] The input layer is used to input the first transaction matrix, the second transaction matrix and the third transaction matrix;

[0038] The feature extraction layer is used to extract features to obtain feature data;

[0039] The fully connected layer is used to process feature data through training weights and training biases;

[0040] The output layer is used to output the supply chain finance risk level.

[0041] Preferably, the expression of the feature extraction layer is:

[0042] ;

[0043] in, is the output of the feature extraction layer, Middle Line The column elements are , Middle Line The column elements are ;

[0044] ;

[0045] ;

[0046] in, The first transaction matrix Middle Line Column elements, The second transaction matrix Middle Line Column elements, is the first elements, is the first elements, , , is the total number of nodes in the supply chain.

[0047] Preferably, the output layer uses a Softmax function to output the supply chain finance risk level.

[0048] The present invention also provides a supply chain finance risk assessment system, which is applied to any one of the supply chain finance risk assessment methods described above, comprising:

[0049] A data acquisition module is used to acquire transaction data on the supply chain within a time period;

[0050] The first matrix establishment module is used to establish the first transaction matrix , the first transaction matrix Line The column elements represent the first Node and The transaction relationship between nodes. The larger the value, the more transactions.

[0051] The second matrix establishment module is used to establish the second transaction matrix , the second transaction matrix Line The column elements represent the first Node and The smaller the value, the closer the relationship between nodes.

[0052] The third matrix establishment module is used to establish the third transaction matrix , the third transaction matrix is ​​a row matrix, whose The element represents the first The risk level of the node, and the smaller the value, the lower the risk;

[0053] A judgment module is used to judge the supply chain financial risk level based on the first transaction matrix, the second transaction matrix and the third transaction matrix through a trained recognition model.

[0054] The technical solution of the embodiment of the present invention has at least the following advantages and beneficial effects:

[0055] The present invention combines big data analysis with neural network-based supply chain financial risk assessment, which can achieve efficient and automated monitoring of financial risk assessment, dynamically capture and analyze changes in the supply chain, and timely identify and respond to potential risks;

[0056] The present invention realizes the extraction of multiple valid data from a large number of risk-related data, establishes multiple characteristic data such as the characteristic conditions of each individual in supply chain finance and the characteristic conditions of the transaction relationship between individuals, which helps to capture more reliable data and data relationships, and assists in high-reliability and accurate risk assessment;

[0057] The present invention improves the security and efficiency of supply chain finance and promotes the healthy development of supply chain finance business;

[0058] The present invention is reasonably designed and easy to implement. It is suitable for evaluation of different products and regional fields, and further suitable for a variety of supply chain finance monitoring scenarios. It has wide applicability and flexible application. BRIEF DESCRIPTION OF THE DRAWINGS

[0059] Figure 1 A flow chart of a supply chain finance risk assessment method provided in Example 1 of the present invention. DETAILED DESCRIPTION

[0060] In order to make the purpose, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Generally, the components of the embodiments of the present invention described and shown in the drawings here can be arranged and designed in various different configurations.

[0061] Example 1

[0062] This embodiment provides a supply chain financial risk assessment method, see Figure 1 , including the following steps:

[0063] Obtain transaction data on the supply chain within a time period;

[0064] Establishing the first trading matrix , the first transaction matrix Line The column elements represent the first Node and The transaction relationship between nodes. The larger the value, the more transactions.

[0065] Establishing the Second Trading Matrix , the second transaction matrix Line The column elements represent the first Node and The smaller the value, the closer the relationship between nodes.

[0066] Establishing the third trading matrix , the third transaction matrix is ​​a row matrix, whose The element represents the first The risk level of the node, and the smaller the value, the lower the risk;

[0067] The supply chain finance risk level is determined based on the first transaction matrix, the second transaction matrix and the third transaction matrix through a trained recognition model.

[0068] This embodiment collects data from multiple aspects, establishes multiple characteristics of each individual in the supply chain financial relationship and the characteristics of the transaction relationship between individuals to capture comprehensive and reliable feature data, and then conducts risk assessment through neural networks. and the second transaction matrix It can reflect the transaction relationship between transaction individuals in the current detection data and past data, the third transaction matrix It can reflect the reliability of each transaction individual. Based on the above data, the supply chain finance relationship can be briefly and comprehensively characterized as feature data, and then risk assessment can be performed through neural network identification in the future. This enables data extraction and construction of reliable and highly accurate supply chain finance risk assessment.

[0069] Example 2

[0070] This embodiment is based on the technical solution of Embodiment 1, and mainly further explains the specific implementation methods of each step of risk assessment. In particular, and Essentially, they represent the first Node and Node, the first Line The elements in the column represent the Node and Some relationship between nodes, so in each matrix and The parameters represented are essentially the same.

[0071] In this embodiment, the method for establishing the first transaction matrix is:

[0072]

[0073] = ;

[0074] in, is the first Line The elements in the column represent the first Node and The number of transactions between nodes, is the first Line The elements in the column represent the first Node and The transaction amount coefficient between nodes, , , and are weights, is the total number of nodes, the matrix and matrix The diagonal elements of are all 0.

[0075] Furthermore, the supply chain Node and The method for obtaining the transaction amount coefficient between nodes is:

[0076] ;

[0077] in, For the supply chain Node and The total amount of transactions between nodes, For the supply chain Node and The total amount of transactions between nodes.

[0078] The first transaction matrix obtained in this embodiment mainly expresses the transaction data relationship between different individuals in the data extracted this time, and specifically includes parameters that can represent the transaction amount and the number of transactions.

[0079] On the other hand, the method for establishing the second transaction matrix is:

[0080] ;

[0081] in, The second transaction matrix Middle Line The elements in the column represent the first Node and The relationship between nodes, , , is the total number of nodes, and the diagonal elements of the second transaction matrix are all 0;

[0082] ;

[0083] in, For the supply chain Node and Historical transaction frequency between nodes, For the supply chain Node and The number of historical transaction product types between nodes, For the supply chain Node and The historical transaction time interval difference parameter between nodes, For the supply chain Node and The average amount of each historical transaction between nodes, , and are weights respectively.

[0084] As a preferred solution of this embodiment, the supply chain Node and The method for obtaining the historical transaction time interval parameters between nodes is:

[0085] Based on historical transactions, the first Node and Continuity between nodes transactions as calculation samples;

[0086] Get :

[0087] ;

[0088] in, Represents the first Node and The number of nodes between Transactions and The time interval between transactions, Represents the first Node and The number of nodes between Transactions and The time interval between transactions.

[0089] Here, the second transaction matrix mainly extracts the transaction relationship between individuals based on the historical transaction data between different individuals, and expresses the transaction closeness between different individuals through the difference in transaction frequency, amount and the time interval between each transaction.

[0090] In addition, the third transaction matrix is ​​established The method is:

[0091] The establishment of the third transaction matrix for:

[0092] ;

[0093] in, The establishment of the third transaction matrix No. elements, representing the first The node's historical transaction parameters, , is the total number of nodes;

[0094] ;

[0095] in, and They are respectively Node close The payment on-time rate and average payment delay time of each transaction, and They are respectively Node close The on-time delivery rate and average delivery delay time of each transaction, For the supply chain Node close The number of violations and defaults in transactions, , and are weights, .

[0096] The third transaction matrix uses historical data to express the credit reliability of each individual in the recent period of time.

[0097] When performing risk assessment, the recognition model preferably includes an input layer, a feature extraction layer, a fully connected layer and an output layer;

[0098] The input layer is used to input the first transaction matrix, the second transaction matrix and the third transaction matrix;

[0099] The feature extraction layer is used to extract features to obtain feature data;

[0100] The fully connected layer is used to process feature data through training weights and training biases;

[0101] The output layer is used to output the supply chain finance risk level.

[0102] As a preferred embodiment, the expression of the feature extraction layer is:

[0103] ;

[0104] in, is the output of the feature extraction layer, Middle Line The column elements are , Middle Line The column elements are ;

[0105] ;

[0106] ;

[0107] in, The first transaction matrix Middle Line Column elements, The second transaction matrix Middle Line Column elements, is the first elements, is the first elements, , , is the total number of nodes in the supply chain.

[0108] This embodiment establishes mathematical expression relationships for two of the three transaction matrices to extract features, which can effectively capture the interactivity between different parameters and improve nonlinearity, which helps the neural network to better learn the relationship correlation behind the parameters during learning and improve the reliability of judgment.

[0109] Finally, the output layer preferably uses a Softmax function to output the supply chain finance risk level. For example, four nodes are set in the output layer and the probabilities of no risk, low risk, medium risk and high risk are output through a Softmax function.

[0110] Example 3

[0111] This embodiment provides a supply chain finance risk assessment system, which is applied to any one of the supply chain finance risk assessment methods described above, including:

[0112] A data acquisition module is used to acquire transaction data on the supply chain within a time period;

[0113] The first matrix establishment module is used to establish the first transaction matrix , the first transaction matrix Line The column elements represent the first Node and The transaction relationship between nodes. The larger the value, the more transactions.

[0114] The second matrix establishment module is used to establish the second transaction matrix , the second transaction matrix Line The column elements represent the first Node and The smaller the value, the closer the relationship between nodes.

[0115] The third matrix establishment module is used to establish the third transaction matrix , the third transaction matrix is ​​a row matrix, whose The element represents the first The risk level of the node, and the smaller the value, the lower the risk;

[0116] A judgment module is used to judge the supply chain financial risk level based on the first transaction matrix, the second transaction matrix and the third transaction matrix through a trained recognition model.

[0117] The above are only preferred embodiments of the present invention and are not intended to limit the present invention. For those skilled in the art, the present invention may have various modifications and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention shall be included in the protection scope of the present invention.

Claims

1. A supply chain finance risk assessment method, characterized in that: The following steps are involved: Obtain transaction data on the supply chain within a time period; Establishing the first trading matrix , the first transaction matrix Line The column elements represent the first Node and The transaction relationship between nodes. The larger the value, the more transactions. Establishing the Second Trading Matrix , the second transaction matrix Line The column elements represent the first Node and The smaller the value, the closer the relationship between nodes. Establishing the third trading matrix , the third transaction matrix is ​​a row matrix, whose The element represents the first The risk level of the node, and the smaller the value, the lower the risk; Determine the supply chain finance risk level through a trained recognition model based on the first transaction matrix, the second transaction matrix, and the third transaction matrix; The method for establishing the first transaction matrix is: = ; in, is the first Line The elements in the column represent the first Node and The number of transactions between nodes, is the first Line The elements in the column represent the first Node and The transaction amount coefficient between nodes, , , and are weights, is the total number of nodes, the matrix and matrix The diagonal elements of are all 0; The supply chain Node and The method for obtaining the transaction amount coefficient between nodes is: ; in, For the supply chain Node and The total amount of transactions between nodes, For the supply chain Node and The total amount of transactions between nodes; The method for establishing the second transaction matrix is: ; in, The second transaction matrix Middle Line The elements in the column represent the first Node and The relationship between nodes, , , is the total number of nodes, and the diagonal elements of the second transaction matrix are all 0; ; in, For the supply chain Node and Historical transaction frequency between nodes, For the supply chain Node and The number of historical transaction product types between nodes, For the supply chain Node and The historical transaction time interval difference parameter between nodes, For the supply chain Node and The average amount of each historical transaction between nodes, , and are weights respectively; The supply chain Node and The method for obtaining the historical transaction time interval parameters between nodes is: Based on historical transactions, the first Node and Continuity between nodes transactions as calculation samples; Get : ; in, Represents the first Node and The number of nodes between Transactions and The time interval between transactions, Represents the first Node and The number of nodes between Transactions and The time interval between transactions; The establishment of the third transaction matrix The method is: The establishment of the third transaction matrix for: ; in, The establishment of the third transaction matrix No. elements, representing the first The node's historical transaction parameters, , is the total number of nodes; ; in, and They are respectively Node close The payment on-time rate and average payment delay time of each transaction, and They are respectively Node close The on-time delivery rate and average delivery delay time of each transaction, For the supply chain Node close The number of violations and defaults in transactions, , and are weights, .

2. A supply chain finance risk assessment method according to claim 1, characterized in that: The recognition model includes an input layer, a feature extraction layer, a fully connected layer and an output layer; The input layer is used to input the first transaction matrix, the second transaction matrix and the third transaction matrix; The feature extraction layer is used to extract features to obtain feature data; The fully connected layer is used to process feature data through training weights and training biases; The output layer is used to output the supply chain finance risk level.

3. A supply chain finance risk assessment method according to claim 2, characterized in that: The expression of the feature extraction layer is: ; in, is the output of the feature extraction layer, Middle Line The column elements are , Middle Line The column elements are ; ; ; in, The first transaction matrix Middle Line Column elements, The second transaction matrix Middle Line Column elements, is the first elements, is the first elements, , , is the total number of nodes in the supply chain.

4. A supply chain finance risk assessment method according to claim 3, characterized in that: The output layer uses a Softmax function to output the supply chain finance risk level.

5. A supply chain finance risk assessment system, applied to a supply chain finance risk assessment method as claimed in any one of claims 1 to 4, characterized in that: include: A data acquisition module is used to acquire transaction data on the supply chain within a time period; The first matrix establishment module is used to establish the first transaction matrix , the first transaction matrix Line The column elements represent the first Node and The transaction relationship between nodes. The larger the value, the more transactions. The second matrix establishment module is used to establish the second transaction matrix , the second transaction matrix Line The column elements represent the first Node and The smaller the value, the closer the relationship between nodes. The third matrix establishment module is used to establish the third transaction matrix , the third transaction matrix is ​​a row matrix, whose The element represents the first The risk level of the node, and the smaller the value, the lower the risk; A judgment module is used to judge the supply chain financial risk level based on the first transaction matrix, the second transaction matrix and the third transaction matrix through a trained recognition model.

Citation Information

Patent Citations

  • Method of controlling financial risks of supply chains

    CN112200486A

  • Steel trade supply chain financial risk assessment method based on fuzzy grey evaluation

    CN113763154A

  • Supply chain financial risk assessment method and system based on big data

    CN114841790A

  • Supply chain financial risk assessment method and device, and medium

    CN115205044A

  • Risk assessment system and method for supply chain, and program product

    CN116307717A