CN117252688A

CN117252688A - Financial risk assessment method, system, terminal equipment and storage medium

Info

Publication number: CN117252688A
Application number: CN202311434516.2A
Authority: CN
Inventors: 陈厚含; 李雯博; 余靓; 林朝曼; 陈侠; 王立彬; 苏雪玲; 陈永华; 叶张雯
Original assignee: Shaanxi Laocaichen Financial Management Co ltd
Current assignee: Shaanxi Laocaichen Financial Management Co ltd
Priority date: 2023-10-31
Filing date: 2023-10-31
Publication date: 2023-12-19

Abstract

The present disclosure relates to the field of financial data analysis, and in particular, to a financial risk assessment method, a system, a terminal device, and a storage medium. The method comprises the steps of carrying out centrality analysis on a risk propagation network, obtaining centrality indexes corresponding to target nodes, and setting key grades corresponding to the target nodes according to the centrality indexes; calibrating corresponding risk initial nodes in the target nodes and risk associated nodes corresponding to the risk initial nodes according to the criticality grades corresponding to the target nodes and the risk associated relations between the target nodes; dividing the risk starting node and the risk associated node into corresponding risk propagation class groups according to the risk assessment grades corresponding to the risk starting node and the risk associated node; and generating a risk assessment report by financial data by combining the risk characteristics and the potential risk diffusion paths corresponding to the risk propagation category group. The technical scheme has the effect of improving the accuracy of financial risk assessment.

Description

Financial risk assessment method, system, terminal equipment and storage medium

Technical Field

The present disclosure relates to the field of financial data analysis, and in particular, to a financial risk assessment method, a system, a terminal device, and a storage medium.

Background

Financial risk assessment refers to the process of assessing and quantifying the potential risk of a financial institution, financial product, or portfolio. It is intended to assist financial institutions and investors in identifying and understanding risks that may be faced and taking appropriate measures to manage and control risks.

Among other things, financial risk assessment generally involves the following aspects: market risk, i.e., the risk of asset value degradation due to market price fluctuations; credit risk, i.e. the risk that the borrower or debtor cannot pay back the debt on time; operational risk, i.e. the risk of loss due to internal process, system or human error; liquidity risk, i.e., the risk that a financial institution or investor cannot obtain sufficient funds in time or render an asset in time; legal risk, i.e., the risk of loss due to factors such as changes in law and regulation, contract disputes, or litigation.

In practical applications, the financial risk assessment usually performs independent assessment on different risks, but ignores the correlation between different risks, and ignoring the correlation may misjudge the overall risk, thereby resulting in reduced accuracy of the financial risk assessment result.

Disclosure of Invention

In order to improve accuracy of financial risk assessment, the application provides a financial risk assessment method, a system, terminal equipment and a storage medium.

In a first aspect, the present application provides a financial risk assessment method, comprising the steps of:

acquiring financial data and associated data among the financial data;

taking the financial data as target nodes, and the associated data as connection between the target nodes to form a corresponding risk propagation network;

performing centrality analysis on the risk propagation network, obtaining a centrality index corresponding to the target node, and setting a criticality grade corresponding to the target node according to the centrality index;

calibrating corresponding risk initial nodes in the target nodes and risk associated nodes corresponding to the risk initial nodes according to the criticality grades corresponding to the target nodes and the risk associated relations between the target nodes;

dividing the risk starting node and the risk associated node into corresponding risk propagation category groups according to the risk assessment grades corresponding to the risk starting node and the risk associated node;

performing path calculation analysis on the risk propagation class group according to a preset risk propagation path algorithm, and generating a potential risk diffusion path corresponding to the risk propagation class group;

And generating a risk assessment report of the financial data by combining the risk characteristics corresponding to the risk propagation category group and the potential risk diffusion path.

By adopting the technical scheme, according to the risk propagation network, the risk propagation paths and the propagation strength among the financial elements can be more intuitively displayed, the risk propagation network is subjected to centrality analysis, centrality indexes such as centrality and betweenness of each target node can be calculated, importance and influence of the nodes in the network can be measured by the indexes, the risk propagation paths and the propagation strength among the financial elements are further divided into corresponding risk propagation category groups according to the criticality grade of the target nodes and the risk association relation among the target nodes, risk assessment analysis and decision of the risk nodes and the reinforced financial data can be better organized and managed, then path calculation analysis is carried out on the risk propagation category groups respectively to generate corresponding potential risk diffusion paths, the risk propagation paths and possible diffusion conditions can be better understood and predicted through the potential risk diffusion paths, and finally corresponding risk assessment reports are generated by combining the risk characteristics corresponding to the risk propagation category groups and the corresponding potential risk diffusion paths among the target nodes in each risk propagation category group. The specific network of various financial data is articulated, and classified evaluation is carried out according to the criticality degree of the target nodes and the risk condition among the target nodes, so that the accuracy of financial risk evaluation is improved.

Optionally, taking the financial data as a target node, and the association data as a connection between the target nodes, forming a corresponding risk propagation network includes the following steps:

acquiring time sequence data corresponding to the target node;

importing the time sequence data into a preset regression model to generate periodic characteristics corresponding to the target node;

modeling the periodic characteristics according to a preset dynamic factor model to generate a corresponding dynamic factor characteristic model;

defining a connection relation corresponding to the associated data according to the characteristic parameters in the dynamic factor characteristic model;

and combining the periodic characteristics corresponding to the target nodes and the connection relation corresponding to the associated data to form the corresponding risk propagation network.

By adopting the technical scheme, the periodic characteristics of the target node are modeled according to the preset dynamic factor model, and the dynamic relation among the financial elements can be captured, so that the interaction and influence among the financial elements can be better obtained, and the authenticity and accuracy of the risk propagation network are improved.

Optionally, defining the connection relationship corresponding to the association data according to the feature parameters in the dynamic factor feature model includes the following steps:

Performing dimension reduction processing on the periodic features according to the dynamic factor feature model to obtain corresponding dynamic factors;

calculating the dynamic factors according to a preset correlation algorithm to obtain corresponding correlation coefficients among the dynamic factors;

and defining the connection relation corresponding to the association data according to the correlation coefficient.

By adopting the technical scheme, the connection relation of the association data between the target nodes is defined according to the corresponding correlation coefficient between the dynamic factors, and a more real risk propagation network can be constructed, so that the accuracy of financial risk assessment can be improved, and a decision maker can better identify and manage risks.

Optionally, performing centrality analysis on the risk propagation network, obtaining a centrality index corresponding to the target node, and setting a criticality grade corresponding to the target node according to the centrality index includes the following steps:

performing centrality analysis on the risk propagation network to generate corresponding centrality analysis data;

carrying out standardization processing on the centrality analysis data according to a preset standardization rule to generate a corresponding standardization interval;

calculating the corresponding standardized data in the standardized interval and the corresponding target weight of the standardized data according to a preset weighting algorithm, and generating a comprehensive centrality index value corresponding to the centrality analysis data;

And setting the criticality grade corresponding to the target node according to the comprehensive centrality index value.

By adopting the technical scheme, the centrality analysis data is standardized according to the preset standardization rule, and centrality indexes of different scales are unified into one standardization interval, so that dimension differences among indexes can be eliminated, and the accuracy of identifying the corresponding key grade of the target node is improved.

Optionally, setting the criticality class corresponding to the target node according to the comprehensive centrality index value includes the following steps:

acquiring the centrality grade corresponding to the target node according to the grade classification standard corresponding to the comprehensive centrality index value;

and evaluating the centrality grade according to a preset grade evaluation rule, and generating a criticality grade corresponding to the target node as the criticality grade.

By adopting the technical scheme, the key grade of the target node is set according to the comprehensive centrality index value, so that the key degree of the node in the risk propagation network can be accurately evaluated and divided, more accurate reference is provided for risk management and decision making, corresponding measures can be taken more pertinently, and the effect and feasibility of risk management are improved.

Optionally, performing path calculation analysis on the risk propagation category group according to a preset risk propagation path algorithm, and generating a potential risk diffusion path corresponding to the risk propagation category group includes the following steps:

setting target allocation weights corresponding to the risk propagation class group by combining preset weight allocation rules and risk propagation characteristics corresponding to the risk propagation class group;

and calculating the target distribution weight according to a preset risk propagation path algorithm, and generating the potential risk diffusion path corresponding to the risk propagation class group.

By adopting the technical scheme, the path calculation analysis is carried out on the target distribution weight of the risk propagation category group according to the preset risk propagation path algorithm, so that the disclosure of the path and the mode of the risk propagation and the better understanding of the mechanism and the trend of the risk propagation are facilitated, and accordingly, the corresponding risk management strategy and measures are effectively formulated, and the diffusion and the influence of potential risks are reduced.

Optionally, after calculating the target allocation weight according to a preset risk propagation path algorithm and generating the potential risk diffusion path corresponding to the risk propagation class group, the method further includes the following steps:

The potential risk diffusion path is led into a preset risk path learning model for training, and a corresponding target risk mode is output;

performing risk assessment on the target risk mode to generate a corresponding risk assessment grade;

and generating a corresponding risk early warning strategy by combining the target risk mode and the risk assessment grade corresponding to the target risk mode.

By adopting the technical scheme, the potential risk diffusion path is converted into the target risk mode, and the risk assessment and the generation of the risk early warning strategy are carried out, so that the potential risks can be identified and handled in advance, the loss caused by the risks is reduced, and the effect of risk management is improved.

In a second aspect, the present application provides a financial risk assessment system comprising:

the data acquisition module is used for acquiring financial data and associated data among the financial data;

the network generation module is used for taking the financial data as a target node, and the associated data are connections among the target nodes to form a corresponding risk propagation network;

the centrality analysis module is used for performing centrality analysis on the risk propagation network, acquiring a centrality index corresponding to the target node, and setting a criticality grade corresponding to the target node according to the centrality index;

The node calibration module is used for calibrating corresponding risk starting nodes in the target nodes and risk association nodes corresponding to the risk starting nodes according to the criticality grades corresponding to the target nodes and the risk association relations between the target nodes;

the risk category classification module is used for classifying the risk starting node and the risk associated node into corresponding risk propagation category groups according to the risk evaluation grades corresponding to the risk starting node and the risk associated node;

the risk diffusion path generation module is used for carrying out path calculation analysis on the risk propagation class group according to a preset risk propagation path algorithm to generate a potential risk diffusion path corresponding to the risk propagation class group;

and the risk assessment module is used for generating a risk assessment report of the financial data by combining the risk characteristics corresponding to the risk propagation category group and the potential risk diffusion path.

By adopting the technical scheme, according to the risk propagation network, the risk propagation paths and the propagation strength among the financial elements can be displayed more intuitively, then the risk propagation network is subjected to centrality analysis through the centrality analysis module, and further centrality indexes such as centrality and betweenness of each target node can be calculated, the importance and influence of the nodes in the network can be measured by the indexes, the risk propagation network is further divided into corresponding risk propagation class groups through the risk class division module according to the criticality class of the target nodes and the risk association relation among the target nodes, risk assessment analysis and decision of the risk nodes and reinforced financial data can be better organized and managed, then the risk propagation class groups are respectively subjected to path calculation analysis through the risk propagation path generation module, corresponding potential risk diffusion paths are generated, the risk propagation paths and possible diffusion conditions can be better understood and predicted through the potential risk diffusion paths, and finally corresponding risk report assessment is generated through the risk assessment module combining the risk characteristics corresponding to the risk propagation class groups and the corresponding potential risk diffusion paths among the target nodes in each risk propagation class group. The specific network of various financial data is articulated, and classified evaluation is carried out according to the criticality degree of the target nodes and the risk condition among the target nodes, so that the accuracy of financial risk evaluation is improved.

In a third aspect, the present application provides a terminal device, which adopts the following technical scheme:

a terminal device comprises a memory and a processor, wherein the memory stores computer instructions capable of running on the processor, and the processor adopts the financial risk assessment method when loading and executing the computer instructions.

By adopting the technical scheme, the computer instruction is generated by the financial risk assessment method and is stored in the memory to be loaded and executed by the processor, so that the terminal equipment is manufactured according to the memory and the processor, and the use is convenient.

In a fourth aspect, the present application provides a computer readable storage medium, which adopts the following technical scheme:

a computer readable storage medium having stored therein computer instructions which, when loaded and executed by a processor, employ a financial risk assessment method as described above.

By adopting the technical scheme, the financial risk assessment method generates the computer instructions, stores the computer instructions in the computer readable storage medium to be loaded and executed by the processor, and facilitates the reading and storage of the computer instructions through the computer readable storage medium.

In summary, the present application includes at least one of the following beneficial technical effects: according to the risk propagation network, the risk propagation paths and the propagation strength among the financial elements can be displayed more intuitively, then the risk propagation network is subjected to centrality analysis, and further centrality indexes such as degree centrality and betweenness centrality of each target node can be calculated, importance and influence of the nodes in the network can be measured by the indexes, the risk propagation network is further divided into corresponding risk propagation category groups according to the criticality level of the target nodes and the risk association relation among the target nodes, risk assessment analysis and decision of the risk nodes and reinforced financial data can be better organized and managed, then the risk propagation category groups are subjected to path calculation analysis respectively to generate corresponding potential risk diffusion paths, the risk propagation paths and possible diffusion conditions can be better understood and predicted through the potential risk diffusion paths, and finally corresponding risk assessment reports are generated by combining the risk characteristics corresponding to the risk propagation category groups and the potential risk diffusion paths corresponding to the target nodes in each risk propagation category group. The specific network of various financial data is articulated, and classified evaluation is carried out according to the criticality degree of the target nodes and the risk condition among the target nodes, so that the accuracy of financial risk evaluation is improved.

Drawings

Fig. 1 is a schematic flow chart of steps S101 to S107 in a financial risk assessment method of the present application.

Fig. 2 is a flowchart illustrating steps S201 to S205 in the financial risk assessment method of the present application.

Fig. 3 is a flowchart illustrating steps S301 to S303 in the financial risk assessment method of the present application.

Fig. 4 is a flowchart illustrating steps S401 to S404 in the financial risk assessment method of the present application.

Fig. 5 is a flowchart illustrating steps S501 to S502 in the financial risk assessment method of the present application.

Fig. 6 is a flowchart illustrating steps S601 to S602 in the financial risk assessment method of the present application.

Fig. 7 is a flowchart illustrating steps S701 to S703 in a financial risk assessment method according to the present application.

FIG. 8 is a block diagram of a financial risk assessment system of the present application.

Reference numerals illustrate:

1. a data acquisition module; 2. a network generation module; 3. a centrality analysis module; 4. a node calibration module; 5. a risk category dividing module; 6. a risk diffusion path generation module; 7. and a risk assessment module.

Detailed Description

The present application is described in further detail below in conjunction with figures 1-8.

The embodiment of the application discloses a financial risk assessment method, as shown in fig. 1, comprising the following steps:

s101, acquiring financial data and associated data among the financial data;

s102, taking financial data as target nodes and associated data as connection between the target nodes to form a corresponding risk propagation network;

s103, carrying out centrality analysis on the risk propagation network, obtaining a centrality index corresponding to the target node, and setting a key grade corresponding to the target node according to the centrality index;

s104, calibrating corresponding risk initial nodes in the target nodes and risk associated nodes corresponding to the risk initial nodes according to the criticality grades corresponding to the target nodes and the risk associated relations among the target nodes;

s105, dividing the risk starting node and the risk associated node into corresponding risk propagation category groups according to risk assessment grades corresponding to the risk starting node and the risk associated node;

s106, carrying out path calculation analysis on the risk propagation class group according to a preset risk propagation path algorithm, and generating a potential risk diffusion path corresponding to the risk propagation class group;

s107, combining the risk characteristics and the potential risk diffusion paths corresponding to the risk propagation category group to generate a risk assessment report of financial data.

In step S101, the financial data refers to data related to a financial market obtained through different data sources and data collection modes, and includes data related to a financial market such as stock price, trading volume, market index, interest rate, exchange rate, and financial statement.

Wherein the financial data may be obtained from a plurality of data sources, including financial institutions, exchanges, data suppliers, financial statements, and the like. Different data sources provide different types and granularities of data, and suitable data sources can be selected according to requirements. Specifically, the collection of the financial data can be performed through an API interface, a data crawler, a data subscription and the like. An API interface is a common way to obtain financial data, and real-time or historical data can be obtained by calling the interface. The data crawler may capture financial data from the web page, and this may be done for data sources without an API interface. The data subscription is to obtain financial data by subscribing to a data service of a data service provider.

Next, the related data is data of a related relationship existing between financial data. Such as an association between stock prices and financial indicators, an association between exchange rates and interest rates, etc. By analyzing and processing the financial data, correlation data, i.e. the degree of correlation and influence between different financial data, can be obtained.

In step S102, risk dissemination refers to how changes and fluctuations between financial data affect other financial data through the transfer of associated data in the risk dissemination network. For example, when stock prices drop, market panic may be raised, resulting in increased trading volume and decreased market index, which in turn affects other financial data.

Specifically, the structure of the risk propagation network refers to a network diagram formed by association relations between financial data. Wherein the network structure may be complex, hierarchical or simple. By analyzing the network structure, the fragile links and key nodes in the financial system can be revealed, and the risk monitoring and management can be facilitated.

In step S103, centrality analysis is a method for measuring importance of nodes in the network. The criticality of the target node in the risk propagation network can be evaluated through the centrality index.

The centrality index is an index for measuring the importance degree of the node in the network. Common centrality indexes include centrality, proximity centrality, median centrality, feature vector centrality and the like. The centrality of degree measures the number of nodes directly connected with other nodes; proximity centrality measures the average distance between a node and other nodes; the medium number centrality measures the frequency of the nodes on the shortest path for connecting other nodes in the network; the feature vector centrality measures the degree of connection of the node with other important nodes.

Secondly, the target nodes can be divided into different criticality grades according to the size of the centrality index. In practical applications, nodes with higher centrality indexes are considered as key nodes, and the influence of the key nodes in the risk propagation network is larger. Nodes can be divided into different key grades such as core nodes, important nodes, secondary nodes and the like according to the size of the centrality index.

It should be noted that centrality analysis helps to identify target nodes with higher criticality in the risk propagation network. The core node is typically the most important node in the overall network, and its variations may have a large impact on the overall network. Important nodes have higher centrality index in the network, but have less impact than core nodes. The centrality index of the secondary nodes in the network is relatively low and its variations have relatively little impact on the overall network.

In step S104, the risk initiating node refers to a target node with a higher level of criticality in the risk propagation network, whose variation may cause risk propagation throughout the network. Through centrality analysis, target nodes with higher levels of criticality, such as core nodes or important nodes, can be identified, and these nodes can be designated as risk initiating nodes.

Secondly, the risk association node refers to other nodes with direct or indirect risk association relation with the risk starting node. There is a risk propagation path between these nodes and risk initiating nodes, whose variation may be affected by the risk initiating nodes. By analyzing the association of risk propagation networks, risk association nodes associated with a particular risk initiation node may be determined.

It should be noted that, the risk initiating node is a key node in the risk propagation network, and its variation may have a major influence on the whole network. By calibrating risk starting nodes, attention and monitoring of the changes of the nodes are facilitated, and potential risk propagation is timely identified and estimated. While risk-related nodes are other nodes that have direct or indirect risk association with the risk-initiating node, their changes may be affected by the risk-initiating node, and thus close attention is also required to the changes of these nodes.

In step S105, the risk initiating node and the risk associating node may be divided into corresponding risk propagation category groups according to the difference in risk assessment level. For example, the high risk category group includes risk start nodes and risk association nodes with a high risk assessment level; the risk category group comprises risk starting nodes and risk association nodes with risk assessment grades of medium risks; the low risk category group includes risk initiating nodes and risk associating nodes with low risk assessment levels.

It should be noted that, according to the classification of risk assessment levels, classifying the risk starting nodes and the risk association nodes into risk propagation category groups is helpful for classifying and managing risks. Nodes in the high risk category group have a higher risk assessment level, whose variations can have a significant risk-spreading impact on the overall network, requiring special attention and management. The node risk assessment level in the risk category group is centered, with changes that have relatively little impact on risk propagation, but still require close attention. The node risk assessment level in the low risk category group is low, and its impact on risk propagation by variations is relatively small, but still requires regular monitoring and management.

In step S106, the preset risk propagation path algorithm is a method or rule set in advance, for determining a risk propagation path between nodes. The algorithm can determine potential risk propagation paths through calculation and analysis based on factors such as association relation, propagation speed, propagation mode and the like among the nodes.

Further, the potential risk diffusion paths calculated and generated for each risk propagation category group through the preset risk propagation path algorithm refer to paths of risks propagated from the initial node to the associated nodes in the risk propagation process. This path may include a plurality of nodes and a plurality of propagation steps, and by analyzing the risk propagation path, the path of the risk propagation and the possible spread may be known.

It should be noted that, the obtained potential risk diffusion paths help to understand the paths and modes of risk propagation in the risk propagation network, identify possible risk propagation paths, and predict the potential risk diffusion range. Corresponding risk management strategies and countermeasures can be formulated at the same time, so that the influence and loss of risk transmission are reduced. Analysis of the risk propagation path also helps to identify potential risk points for risk propagation, and preventive measures are taken in advance to reduce risk likelihood.

In step S107, risk characteristics refer to characteristics and attributes of the risk propagation category group. Depending on the different risk assessment levels and risk propagation paths, the risk characteristics may include the degree of risk, the source of the risk, the risk propagation speed, the complexity of the risk propagation path, etc. By analyzing the risk characteristics, the nature and the characteristics of the risk can be known.

Secondly, a risk assessment report is a report for assessing and analyzing risk, and according to analysis of risk characteristics and potential risk diffusion paths, a risk assessment report for financial data can be generated. The report may include the extent of potential impact of the risk, the extent of possible loss, the route and speed of risk propagation, the potential risk points of risk propagation, and the like.

Specifically, the generation of the risk assessment report can be roughly divided into the following steps: risk feature analysis corresponding to the risk propagation category group: first, risk profile analysis is performed for each risk propagation category group. This includes assessing the risk level, risk source, risk propagation speed, complexity of the risk propagation path, etc. for each category group. These features can be described using statistics or historical cases.

Second, risk potential diffusion path analysis: based on a preset risk propagation path algorithm, for each risk propagation class group, a potential risk propagation path is determined and corresponding risk characteristics are matched. This can be done by analyzing the association, propagation rate and propagation manner between nodes in the risk propagation class group. Network analysis, timing analysis, etc. methods may be used to determine potential risk diffusion paths.

Further, according to the obtained risk characteristics and the analysis result of the potential risk diffusion path, a risk assessment report is written. Specifically, the content includes a risk profile: summarizing and summarizing risk characteristics of the risk propagation class group, including risk degrees, risk sources and the like; potential risk diffusion path: giving a corresponding potential risk diffusion path for each risk propagation class group, and describing a path for risk propagation from the initial node to the associated node; risk assessment results: according to the analysis of the risk propagation path and the risk characteristics, the potential influence range, the possible loss degree and the like of the risk are evaluated; advice and measures: based on the risk assessment results, corresponding suggestions and measures are provided to help financial institutions or related departments reduce the risk potential and impact.

It should be noted that, the risk assessment report may be verified and updated according to the actual situation. By comparing and feeding back with actual risk events, the risk assessment report is continuously perfected and updated so as to improve the accuracy and reliability of assessment.

According to the financial risk assessment method provided by the embodiment, according to the risk propagation network, the risk propagation paths and the propagation strength among financial elements can be displayed more intuitively, then the risk propagation network is subjected to centrality analysis, so that centrality indexes such as centrality and betweenness of each target node can be calculated, importance and influence of the nodes in the network can be measured by the indexes, the risk propagation paths and the propagation strength among the target nodes are further divided into corresponding risk propagation category groups according to the criticality grade of the target nodes and the risk association relation among the target nodes, risk assessment analysis and decision of the risk nodes and the reinforced financial data can be better organized and managed, then path calculation analysis is carried out on the risk propagation category groups respectively, corresponding potential risk diffusion paths are generated, the risk propagation paths and possible diffusion conditions can be better understood and predicted through the potential risk diffusion paths, and finally corresponding risk assessment reports are generated by combining the risk characteristics corresponding to the risk propagation category groups and the corresponding potential risk diffusion paths among the target nodes in each risk propagation category group. The specific network of various financial data is articulated, and classified evaluation is carried out according to the criticality degree of the target nodes and the risk condition among the target nodes, so that the accuracy of financial risk evaluation is improved.

In one implementation manner of the present embodiment, as shown in fig. 2, step S102, that is, taking financial data as a target node and associated data as a connection between target nodes, forms a corresponding risk propagation network, includes the following steps:

s201, obtaining time sequence data corresponding to a target node;

s202, importing the time sequence data into a preset regression model to generate periodic characteristics corresponding to the target node;

s203, modeling periodic characteristics according to a preset dynamic factor model, and generating a corresponding dynamic factor characteristic model;

s204, defining a connection relation corresponding to the associated data according to the characteristic parameters in the dynamic factor characteristic model;

s205, combining the periodic characteristics corresponding to the target nodes and the connection relations corresponding to the associated data to form a corresponding risk propagation network.

In steps S201 to S202, the time-series data refers to a collection of data points collected for a specific target node (e.g., economic index, financial market index, stock price, etc.) during a period of time. The data points are arranged in a time sequence to form a time series. Each data point represents the value of the target node at a particular point in time.

Second, the preset regression model is a statistical model for analyzing the relationship between variables. The relationship between the independent variable and the dependent variable is described by establishing a mathematical equation, and the value of the dependent variable is predicted by using the known independent variable value. Regression models may be used to interpret and predict data, reveal correlations between variables, and make causal inferences. Specifically, the time series data is imported into a preset regression model, so that periodic characteristics corresponding to the target node can be generated. The preset regression model may be a linear regression model, an ARIMA model, a GARCH model, and the like, and the specific selection of which model depends on the nature of the data and the purpose of analysis.

In the regression model, the periodic characteristics may be captured by introducing appropriate arguments. Common methods include introducing seasonal variables, time trend variables, or periodic indicators, etc. These variables can help the model capture periodic fluctuations and trends in the data. The periodic features can reveal periodic fluctuations and seasonal variations in the data, which can help understand the periodicity laws of the data and predict future fluctuations. By interpreting the periodic features, the user can be aided in understanding the economic or financial reasons behind the data, as well as the factors that affect the variability of the data.

In steps S203 to S204, the preset dynamic factor model is a statistical model for analyzing time-series data, which combines the dynamic factor analysis and the time-series modeling methods. The model assumes that the observed data is composed of potential dynamic factors and observation errors, and describes the dynamic change and the observation relation of the data by establishing a state equation and an observation equation of the dynamic factors.

Wherein, in the preset dynamic factor model, the observed data is represented as a linear combination of the dynamic factor and the observed error. The dynamic factors represent common patterns of variation that are not observed in the data and can be modeled by state equations. The state equations are typically represented using an Autoregressive (AR) model or a Vector Autoregressive (VAR) model to capture the time dynamics and correlation of the dynamic factors. By presetting the dynamic factor model, the state equation parameters of the dynamic factor, the observation matrix and the variance of the observation error can be estimated, and fitting and prediction of the model can be performed. The model can be used for analyzing common variation, correlation and predictive performance of time series data, revealing potential modes and structures in the data, thereby providing more accurate and reliable data analysis results and decision basis.

In particular, the periodic features are modeled according to a preset dynamic factor model, and a method of dynamic factor analysis (Dynamic Factor Analysis) may be used. Dynamic factor analysis is a multivariate statistical model used to analyze and predict dynamic changes of multiple time series.

Next, according to the dynamic factor model described above, it is possible to model by expressing the observed data, i.e., the periodic characteristics, as a linear combination of the dynamic factor and the observed error. The model can be expressed as: x (t) =Λf (t) +epsilon (t). Where X (t) is the vector of the observed data, Λ is the observation matrix, F (t) is the vector of the dynamic factor, ε (t) is the vector of the observation error. The observation matrix Λ represents a linear relationship between the dynamic factor and the observed data. The change in the dynamic factor F (t) may be modeled by a state equation, and may be represented using an Autoregressive (AR) model or a Vector Autoregressive (VAR) model.

Further, according to the feature parameters in the dynamic factor feature model, a connection relationship between the associated data may be defined. The characteristic parameters may include an observation matrix Λ, state equation parameters of the dynamic factors, variances of the observation errors, and the like. These parameters define the relationship between the dynamic factors in the data and the observed errors, as well as the correlation between the observed data. The connection relationship may be defined by an observation matrix Λ. Each row of the observation matrix Λ represents a linear relationship between the feature corresponding to the dynamic factor and the observation data. The degree of contribution of each dynamic factor to the observed data can be determined by observing the matrix Λ, thereby revealing the degree of correlation and influence between different features.

In step S205, the target node is represented as a node in the network, which may represent a risk or source of risk of interest. And then, determining the connection strength between other nodes and the target node according to the connection relation of the associated data. These connections may be direct or indirect, reflecting the relationships and propagation capabilities between different nodes.

Second, the periodic characteristics of the target node are utilized to determine a temporal pattern of risk propagation. For example, if the periodic characteristics of the target node appear to be seasonal variations, the nodes and connections of the network may be mapped to seasons to reflect the effects of the seasons on risk spread.

The influence degree of different nodes on the target node can be evaluated through the connection relation and the node characteristics. The connection relationship represents the relationship and propagation capability between nodes, and the node characteristics reflect the attribute of the node itself. By taking these factors into account, it can be determined which nodes have a greater impact on the risk propagation of the target node. By associating the periodic characteristics of the target node with the network node and the connection relationship, the time pattern of risk propagation can be analyzed. The risk management method and the risk management system are beneficial to understanding the change and fluctuation of risk propagation in different time periods, and provide more accurate information for risk management and decision making.

According to the financial risk assessment method provided by the embodiment, the periodic characteristics of the target nodes are modeled according to the preset dynamic factor model, so that the dynamic relation among financial elements can be captured, the interaction and influence among the financial elements can be better obtained, and the authenticity and accuracy of a risk propagation network are improved.

In one implementation manner of the present embodiment, as shown in fig. 3, step S204, that is, defining the connection relationship corresponding to the association data according to the feature parameters in the dynamic factor feature model, includes the following steps:

s301, performing dimension reduction processing on periodic features according to a dynamic factor feature model to obtain corresponding dynamic factors;

s302, calculating dynamic factors according to a preset correlation algorithm, and obtaining corresponding correlation coefficients among the dynamic factors;

s303, defining a connection relation corresponding to the associated data according to the correlation coefficient.

In step S301, the periodic feature of the target node may be subjected to a dimension reduction process by using the dynamic factor feature model. I.e. a dynamic factor is extracted from the original periodic features to represent the dominant pattern of variations in the data. The dynamic factor is a factor that can explain a main pattern or direction of data fluctuation in time-series data. The method is obtained by performing dimension reduction processing on the original data, and can represent main variation characteristics in the data.

In particular, in time series data, there are typically a number of variables or features, and there may be some correlation between these features. The purpose of the dynamic factor is to extract key patterns or directions that can explain the data changes by comprehensively analyzing the relevant features. Wherein the dynamic factor may be extracted by some mathematical method, such as Principal Component Analysis (PCA) or factor analysis. These methods can transform the raw feature data into a new set of synthesis factors, which are linear combinations of the raw features. Moreover, these new factors are ordered in terms of importance in interpreting the data changes, so that selecting the first few factors can better interpret the changes in the original data.

In steps S302 to S303, the preset correlation algorithm refers to a specific algorithm or method used in calculating the correlation coefficient. Such as pearson correlation coefficients and spearman scale correlation coefficients.

Wherein, according to a preset correlation algorithm, the correlation coefficient between dynamic factors can be calculated. The correlation coefficient is used for measuring the linear correlation degree between two variables, so that the correlation relationship between dynamic factors can be known.

And secondly, after the correlation coefficient between the dynamic factors is calculated, the connection relation between the associated data can be defined. The connection relationship may be determined by setting a threshold value, and when the value of the correlation coefficient exceeds the threshold value, it may be considered that there is a connection relationship between the two dynamic factors.

Specifically, according to a preset correlation algorithm (such as pearson correlation coefficient or spearman class correlation coefficient), the dynamic factors are calculated to obtain correlation coefficients between the dynamic factors, and then, according to the value of the correlation coefficients, a threshold value can be set to define the connection relationship. When the absolute value of the correlation coefficient is greater than or equal to the threshold value, the connection relationship between the two dynamic factors is considered to exist. A network graph of connection relationships may then be created, wherein each dynamic factor may represent a node in the network and the connection relationships represent edges in the network.

According to the financial risk assessment method provided by the embodiment, the connection relation of the association data between the target nodes is defined according to the corresponding correlation coefficient between the dynamic factors, and a more real risk propagation network can be constructed, so that the accuracy of financial risk assessment can be improved, and a decision maker can better identify and manage risks.

In one implementation manner of the present embodiment, as shown in fig. 4, step S103 of performing a centrality analysis on the risk propagation network to obtain a centrality index corresponding to the target node, and setting a criticality class corresponding to the target node according to the centrality index includes the following steps:

s401, carrying out centrality analysis on the risk propagation network to generate corresponding centrality analysis data;

s402, carrying out standardization processing on the centrality analysis data according to a preset standardization rule to generate a corresponding standardization interval;

s403, calculating the standardized data corresponding to the standardized interval and the target weight corresponding to the standardized data according to a preset weighting algorithm, and generating a comprehensive centrality index value corresponding to centrality analysis data;

s404, setting a criticality grade corresponding to the corresponding target node according to the comprehensive centrality index value.

In step S401, the centrality analysis data is result data obtained by performing centrality analysis on each node in the network. These data are used to describe and measure the importance and impact of each node in the network, as well as the location and connectivity of the nodes in the network. The centrality analysis data comprise indexes such as centrality, approaching centrality, medium centrality, characteristic vector centrality and the like. For each node, its value on the respective centrality index may be calculated.

For example, for degree centrality, the degree of each node, i.e., the number of edges directly connected to the node, may be recorded. For near centrality, the average shortest path length between each node and other nodes may be calculated. For the betting centrality, the number of occurrences of each node in all shortest paths in the network can be calculated. For feature vector centrality, the degree of connection of each node to other important nodes can be calculated.

It should be noted that the centrality analysis data provides a quantitative indicator of the importance of the nodes, which can be used to compare and rank the nodes in the network. By analyzing the centrality analysis data, key nodes and core nodes in the network can be identified, and influence and control force of the nodes are known.

In step S402, the preset normalization rule refers to a rule or method preset at the time of performing data normalization processing. Normalization is a common data preprocessing technique used to convert raw data into a unified, standardized form to facilitate comparison and analysis of differences between different metrics or data.

Furthermore, according to a preset standardization rule, the standardization processing can be performed on the centrality analysis data, and the data is mapped to a unified standardization interval. Normalization may allow for comparability of different indicators for better comparison and analysis.

In particular, raw data, referred to herein as centrality analysis data, may be mapped linearly to a specified normalized interval using a min-max normalization method. The method can be calculated by the following formula: normalized value = (original value-minimum)/(maximum-minimum) × (max interval-min interval) +min interval. The minimum value and the maximum value are respectively the minimum value and the maximum value in the original data, and the maximum interval and the minimum interval are the upper limit and the lower limit of a preset standardized interval.

In step S403 to step S404, the preset weighting algorithm refers to an algorithm for performing weighting calculation on different centrality indexes in the centrality analysis. In centrality analysis, a plurality of centrality indices are typically considered, each index corresponding to a feature or attribute in the dataset, for evaluating the importance and impact of a node. The purpose of the preset weighting algorithm is to assign different weights to different centrality indexes according to the demands of users or analyzers so as to reflect the relative importance of the centrality indexes in comprehensive centrality index calculation. The contribution of different indexes can be more accurately comprehensively considered through weighted calculation, so that a more comprehensive and accurate centrality analysis result is obtained.

For example, the preset weighting algorithm is an analytic hierarchy process, and according to the relative importance of different indexes, the proportional relation of the weights is obtained by constructing a judgment matrix and calculating the characteristic values. The method is characterized in that the problems are layered, and the importance degree of each index relative to other indexes is determined through comparison and judgment, so that the relative proportion of the weights is obtained.

The standardized data corresponding to the standardized interval refers to data which converts centrality analysis data into data of relative proportion or relative position, so that comparison and comprehensive analysis can be performed between different indexes or variables. The purpose of the standardized data is to eliminate dimensional differences between different indexes or variables, convert them into a unified scale, and facilitate comprehensive analysis and comparison.

Secondly, the target weight corresponding to the standardized data refers to that a weight value is allocated to each standardized data according to the relative importance or priority of the data, and the weight value is used for carrying out weight calculation on the data in a weight algorithm. The target weight reflects the relative importance or contribution of each normalized data in the overall calculation. Through data analysis and statistical methods, weight calculation can be performed according to indexes such as distribution, variance, correlation and the like of data. For example, the weights are determined from variance contribution ratios or factor loadings of the data using Principal Component Analysis (PCA) or factor analysis, or the like.

Further, assuming that there are n centrality indices, the corresponding normalized data is x1, x2,..xn, and the corresponding target weights are w1, w2,..wn. The integrated centrality index value may be calculated by means of weighted summation: composite centrality index value = w1×1+w2×2+ + wn×n. Where wi represents the weight of the i-th index and xi represents the normalized data of the i-th index. According to the comprehensive centrality index value, the criticality grade of the corresponding target node can be set.

In practical applications, the criticality class may be partitioned according to the size of the composite centrality index value. For example, the division modes include high, medium, and low level division: namely, the nodes are classified into high, medium and low levels according to the magnitude of the comprehensive centrality index value, for example, when the comprehensive centrality index value is greater than a certain threshold value, the node is set to be high, when the comprehensive centrality index value is between certain threshold values, the node is set to be medium, and when the comprehensive centrality index value is less than a certain threshold value, the node is set to be low; dividing a numerical range: that is, according to the magnitude of the comprehensive centrality index value, the nodes are divided into different numerical ranges, for example, the index values are divided into ranges of 0-10, 11-20, 21-30, and the like, and each range corresponds to a different criticality grade.

According to the financial risk assessment method provided by the embodiment, the centrality analysis data is standardized according to the preset standardized rule, and centrality indexes of different scales are unified into one standardized interval, so that dimensional differences among indexes can be eliminated, and accuracy of identification of the target node corresponding to the key grade is improved.

In one implementation manner of this embodiment, as shown in fig. 5, step S404, that is, setting the criticality level corresponding to the corresponding target node according to the comprehensive centrality index value includes the following steps:

s501, acquiring the centrality grade of a corresponding target node according to grade division standards corresponding to the comprehensive centrality index values;

s502, evaluating the centrality grade according to a preset grade evaluation rule, and generating a criticality grade corresponding to the target node as the criticality grade.

In steps S501 to S502, the ranking criterion is a set of rules or criteria for classifying the comprehensive centrality index value of the target node into different ranks. These criteria are typically based on the distribution of comprehensive centrality index values and specific application requirements for classifying and assessing the centrality level of the target node. The centrality level refers to the comprehensive centrality index value of the target node, which is divided into different levels or categories.

For example, the index value range may be classified into several different levels, such as very low, medium, high, very high, etc., according to a level classification criterion of the integrated centrality index value. The partitioning criteria may be determined according to specific data distribution conditions and requirements, so as to ensure that the partitioning result can reflect the centrality degree of the target node in the network.

Next, the preset rating rule refers to a set of rules or criteria set in advance when the centrality rating is divided. These rules or criteria are criteria for mapping the aggregate centrality index value of the target node to a corresponding level, determined in advance, based on the particular needs and data distribution.

The preset grade evaluation rule can be determined based on different index value ranges and distribution conditions, so that the importance and influence of the target node in the network can be accurately reflected by the centrality grade. The content includes the index value range: that is, upper and lower limits of index values of different levels are defined to determine the range of the level division; number of grades: i.e. how many different levels to divide into, typically according to the data distribution and requirements; grade label: i.e. each level is given a corresponding label or name for ease of understanding and application; the grading mode is as follows: it is determined by what method the index value is mapped to the corresponding level, for example, using equidistant division or the like.

Further, comparing the obtained centrality grade of the target node with a preset grade evaluation rule, and determining the criticality grade of the target node. For example, if the centrality level of the target node is very high, its criticality level may be rated very high according to a preset rating rule.

In the financial risk assessment method provided in this embodiment, as shown in fig. 6, step S106 is to perform path calculation analysis on the risk propagation class group according to a preset risk propagation path algorithm, and generating a potential risk diffusion path corresponding to the risk propagation class group includes the following steps:

s601, setting target allocation weights of corresponding risk propagation category groups by combining preset weight allocation rules and risk propagation characteristics corresponding to the risk propagation category groups;

s602, calculating target distribution weights according to a preset risk propagation path algorithm, and generating potential risk diffusion paths corresponding to the risk propagation class groups.

In steps S601 to S602, the preset weight allocation rule refers to a preset set of rules or methods for allocating weights to different target nodes when determining the risk propagation path or evaluating the importance of the target nodes. These rules or methods are criteria for determining the importance or contribution of a target node in the risk propagation process, determined in advance, based on specific needs and considerations.

Secondly, the risk propagation characteristics corresponding to the risk propagation category group refer to risk propagation characteristics or characteristics with commonality and similarity in different risk propagation categories or groups. These features describe some common laws, patterns, or trends of risk propagation in that class or group. For example, the risk propagation features include features of propagation paths, propagation ranges, and the like.

Specifically, according to a preset weight distribution rule and a risk propagation characteristic corresponding to a risk propagation class group, a target distribution weight corresponding to the risk propagation class group can be set. These weights are used to represent the importance or contribution of different target nodes during the risk propagation process. By setting appropriate target allocation weights, the potential risk diffusion path and the impact of risk propagation can be more accurately assessed.

Further, the preset risk propagation path algorithm is an algorithm for determining a risk propagation path, which identifies and analyzes a path of risk propagation and a path influencing factor based on preset rules and methods. The algorithm can simulate a potential risk propagation path, namely a potential risk diffusion path according to a network topological structure, centrality of target nodes and weight distribution conditions. The algorithm can calculate the potential risk diffusion path and the influence degree of the propagation path according to different weight distribution and connection relations among the nodes.

The risk potential diffusion path refers to a path and factors possibly causing risk diffusion in the risk management and analysis process through analysis and prediction of the propagation path of the risk potential. From this risk potential propagation path, the risk potential can be identified, analyzed and evaluated, as well as inference and prediction of the risk propagation path.

According to the financial risk assessment method provided by the embodiment, path calculation analysis is carried out on the target distribution weight of the risk propagation class group according to the preset risk propagation path algorithm, so that the disclosure of the path and mode of risk propagation and better understanding of the mechanism and trend of risk propagation are facilitated, corresponding risk management strategies and measures are effectively formulated, and the diffusion and influence of potential risks are reduced.

In one implementation manner of this embodiment, as shown in fig. 7, in step S106, the calculation of the target allocation weight according to the preset risk propagation path algorithm, and the generation of the potential risk diffusion path corresponding to the risk propagation class group further include the following steps:

s701, importing a potential risk diffusion path into a preset risk path learning model for training, and outputting a corresponding target risk mode;

S702, performing risk assessment on the target risk mode to generate a corresponding risk assessment grade;

s703, combining the target risk mode and the risk assessment grade corresponding to the target risk mode to generate a corresponding risk early warning strategy.

In steps S701 to S702, the preset risk path learning model is a model trained based on known risk propagation rules and factors, which can learn and identify potential risk propagation paths and predict possible target risk patterns. The target risk mode is a risk propagation path mode obtained after training of a learning model on the basis of the potential risk diffusion path. Such patterns can be used to describe changes and evolution of potential risks that may occur during propagation, including changes in risk sources, propagation pathways, impact ranges, and the like.

Further, risk assessment is performed on the target risk pattern, and a corresponding risk assessment level can be generated. Risk assessment is the determination of the severity and possible impact range of a potential risk by analyzing and assessing the target risk pattern. According to the evaluation result, the risk can be classified into different grades, such as high risk, medium risk, low risk, etc., for example, so that decisions on risk management and control can be made.

In step S703, the risk early warning policy is a series of measures and methods formulated to discover and cope with the potential risk in time. Based on the analysis results of the target risk mode and the risk assessment level, the risk event possibly happening is early warned in advance, and corresponding measures are taken to reduce the occurrence and influence of the risk.

Specifically, the risk early warning strategy includes early warning criteria: determining early warning standards of different grades according to the risk assessment grade, wherein the early warning standards are more strict for high risk grades, and more sensitive and timely early warning measures are needed; while for low risk levels, the pre-warning criteria may be relatively relaxed; early warning mode: determining potential risk early warning modes and channels, which can include real-time monitoring by using sensors and monitoring equipment, risk early warning by using a data analysis and prediction model, early warning information sending to related personnel through a communication and information transmission system and the like; early warning level: determining early warning levels of different levels according to the risk assessment level, for example, for a high risk level, an emergency early warning level may need to be adopted, and emergency response measures are immediately started; for low risk levels, the early warning level may be relatively low, and more conventional early warning measures may be taken; emergency response: and according to the risk assessment level, corresponding emergency response measures and strategies are formulated.

According to the financial risk assessment method, the potential risk diffusion path is converted into the target risk mode, and risk assessment and risk early warning strategies are generated, so that potential risks can be identified and handled in advance, loss caused by the risks is reduced, and the effect of risk management is improved.

The embodiment of the application discloses a financial risk assessment system, as shown in fig. 8, including:

the data acquisition module 1 is used for acquiring financial data and associated data among the financial data;

the network generation module 2 is used for forming a corresponding risk propagation network by taking financial data as target nodes and associated data as connection between the target nodes;

the centrality analysis module 3 is used for performing centrality analysis on the risk propagation network, obtaining a centrality index corresponding to the target node, and setting a key grade corresponding to the target node according to the centrality index;

the node calibration module 4 is used for calibrating corresponding risk initial nodes in the target nodes and risk associated nodes corresponding to the risk initial nodes according to the criticality grades corresponding to the target nodes and the risk associated relations among the target nodes;

the risk category classification module 5 is configured to classify the risk starting node and the risk associated node into corresponding risk propagation category groups according to risk assessment levels corresponding to the risk starting node and the risk associated node;

The risk diffusion path generation module 6 performs path calculation analysis on the risk propagation class group according to a preset risk propagation path algorithm to generate a potential risk diffusion path corresponding to the risk propagation class group;

and the risk assessment module 7 is used for generating a risk assessment report of the financial data by combining the risk characteristics and the potential risk diffusion paths corresponding to the risk propagation category group.

According to the financial risk assessment system provided by the embodiment, according to the risk propagation network, the risk propagation paths and the propagation strength among the financial elements can be displayed more intuitively, then the risk propagation network is subjected to centrality analysis through the centrality analysis module 3, so that centrality indexes such as centrality and betweenness centrality of each target node can be calculated, importance and influence of the nodes in the network can be measured, the risk propagation network is further divided into corresponding risk propagation category groups through the risk category division module 5 according to the criticality grades of the target nodes and the risk association relations among the target nodes, risk assessment analysis and decision of the risk nodes and reinforced financial data can be better organized and managed, then path calculation analysis is performed on the risk propagation category groups through the risk diffusion path generation module 6 to generate corresponding potential risk diffusion paths, the risk propagation paths and possible diffusion conditions can be better understood and predicted through the potential risk diffusion paths, and finally the risk report corresponding risk report assessment is generated through the risk assessment module 7 combining the corresponding risk characteristics of the risk propagation category groups and the potential risk diffusion paths corresponding to the target nodes in the risk propagation category groups. The specific network of various financial data is articulated, and classified evaluation is carried out according to the criticality degree of the target nodes and the risk condition among the target nodes, so that the accuracy of financial risk evaluation is improved.

It should be noted that, the financial risk assessment system provided in the embodiment of the present application further includes each module and/or corresponding sub-module corresponding to the logic function or logic step of any one of the foregoing financial risk assessment methods, so that the same effects as each logic function or logic step are achieved, and specifically will not be described herein.

The embodiment of the application also discloses a terminal device, which comprises a memory, a processor and computer instructions stored in the memory and capable of running on the processor, wherein when the processor executes the computer instructions, any one of the financial risk assessment methods in the embodiment is adopted.

The terminal device may be a computer device such as a desktop computer, a notebook computer, or a cloud server, and the terminal device includes, but is not limited to, a processor and a memory, for example, the terminal device may further include an input/output device, a network access device, a bus, and the like.

The processor may be a Central Processing Unit (CPU), or of course, according to actual use, other general purpose processors, digital Signal Processors (DSP), application Specific Integrated Circuits (ASIC), ready-made programmable gate arrays (FPGA) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc., and the general purpose processor may be a microprocessor or any conventional processor, etc., which is not limited in this application.

The memory may be an internal storage unit of the terminal device, for example, a hard disk or a memory of the terminal device, or may be an external storage device of the terminal device, for example, a plug-in hard disk, a Smart Memory Card (SMC), a secure digital card (SD), or a flash memory card (FC) provided on the terminal device, or the like, and may be a combination of the internal storage unit of the terminal device and the external storage device, where the memory is used to store computer instructions and other instructions and data required by the terminal device, and the memory may be used to temporarily store data that has been output or is to be output, which is not limited in this application.

Any one of the financial risk assessment methods in the embodiments is stored in the memory of the terminal device through the terminal device, and is loaded and executed on the processor of the terminal device, so that the terminal device is convenient to use.

The embodiment of the application also discloses a computer readable storage medium, and the computer readable storage medium stores computer instructions, wherein when the computer instructions are executed by a processor, any one of the financial risk assessment methods in the above embodiments is adopted.

The computer instructions may be stored in a computer readable medium, where the computer instructions include computer instruction codes, where the computer instruction codes may be in a source code form, an object code form, an executable file form, or some middleware form, etc., and the computer readable medium includes any entity or device capable of carrying the computer instruction codes, a recording medium, a usb disk, a mobile hard disk, a magnetic disk, an optical disk, a computer memory, a read-only memory (ROM), a Random Access Memory (RAM), an electrical carrier signal, a telecommunication signal, a software distribution medium, etc., where the computer readable medium includes but is not limited to the above components.

Any one of the financial risk assessment methods in the above embodiments is stored in the computer readable storage medium through the present computer readable storage medium, and is loaded and executed on a processor, so as to facilitate storage and application of the method.

The foregoing are all preferred embodiments of the present application, and are not intended to limit the scope of the present application in any way, therefore: all equivalent changes in structure, shape and principle of this application should be covered in the protection scope of this application.

Claims

1. A financial risk assessment method, comprising the steps of:

acquiring financial data and associated data among the financial data;

2. The financial risk assessment method according to claim 1, wherein the financial data is used as a target node, the associated data is used as a connection between the target nodes, and the forming of the corresponding risk propagation network includes the following steps:

acquiring time sequence data corresponding to the target node;

3. The financial risk assessment method according to claim 2, wherein defining the connection relationship corresponding to the association data according to the feature parameters in the dynamic factor feature model comprises the steps of:

4. The financial risk assessment method according to claim 1, wherein performing a centrality analysis on the risk propagation network, obtaining a centrality index corresponding to the target node, and setting a criticality class corresponding to the target node according to the centrality index comprises the steps of:

5. The financial risk assessment method according to claim 4, wherein setting the criticality class corresponding to the target node according to the integrated centrality index value comprises the steps of:

6. The financial risk assessment method according to claim 1, wherein the performing path computation analysis on the risk propagation class group according to a preset risk propagation path algorithm, and generating a potential risk diffusion path corresponding to the risk propagation class group includes the following steps:

7. The financial risk assessment method according to claim 1, further comprising the steps of, after calculating the target allocation weights according to a preset risk propagation path algorithm, generating the potential risk diffusion paths corresponding to the risk propagation category group:

8. A financial risk assessment system, comprising:

a data acquisition module (1) for acquiring financial data and associated data between the financial data;

the network generation module (2) is used for forming a corresponding risk propagation network by taking the financial data as a target node and the associated data as connection between the target nodes;

the centrality analysis module (3) is used for performing centrality analysis on the risk propagation network, obtaining a centrality index corresponding to the target node, and setting a criticality grade corresponding to the target node according to the centrality index;

the node calibration module (4) is used for calibrating corresponding risk initial nodes in the target nodes and risk associated nodes corresponding to the risk initial nodes according to the criticality grades corresponding to the target nodes and the risk associated relations between the target nodes;

The risk category classification module (5) is used for classifying the risk starting node and the risk associated node into corresponding risk propagation category groups according to the risk evaluation grades corresponding to the risk starting node and the risk associated node;

the risk diffusion path generation module (6) is used for carrying out path calculation analysis on the risk propagation class group according to a preset risk propagation path algorithm to generate a potential risk diffusion path corresponding to the risk propagation class group;

and the risk assessment module (7) is used for generating a risk assessment report of the financial data by combining the risk characteristics corresponding to the risk propagation category group and the potential risk diffusion path.

9. A terminal device comprising a memory and a processor, wherein the memory has stored therein computer instructions executable on the processor, which processor, when loaded and executed, employs a financial risk assessment method according to any one of claims 1 to 7.

10. A computer readable storage medium having stored therein computer instructions which, when loaded and executed by a processor, employ a financial risk assessment method according to any one of claims 1 to 7.