Financial risk prediction method and system based on big data

Through the collation, analysis and classification marking of financial market data, the problem of uneven data quality value has been solved, the accuracy of risk prediction and the risk management capabilities of financial institutions have been improved, systemic risks have been reduced, and more accurate risk warning has been achieved. and loss reduction.

CN118941391AInactive Publication Date: 2024-11-12JIANGSU CHAOLI ELECTRIC
View PDF 0 Cites 1 Cited by

Patent Information

Application Number
CN202410465370.6
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-04-18
Publication Date
2024-11-12
Estimated Expiration
Not applicable · inactive patent

AI Technical Summary

Technical Problem

In the financial market, as the amount of data increases, the value of data quality becomes uneven, and existing systems find it difficult to identify high-quality data in a timely manner, resulting in insufficient risk management and increasing the possibility of systemic risks and losses.

Method used

By collecting large-scale data related to the financial market, sorting and cleaning, exploratory analysis and real-time market analysis, evaluating the trend correlation and impact degree between data, comprehensively analyzing and evaluating the quality value of data, and classifying and labeling the data. Warn of potential risks in advance.

Benefits of technology

It improves the value of data quality, enhances the accuracy of risk prediction and the risk perception ability of financial institutions, reduces systemic risks, promptly warns of risk events, and helps financial institutions take countermeasures and reduces losses.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118941391A_ABST
    Figure CN118941391A_ABST
Patent Text Reader

Abstract

The invention discloses a financial risk prediction method and system based on big data, and particularly relates to the technical field of data analysis. The method comprises the following steps of: collecting various data from a financial market by utilizing a big data technology and an algorithm, including information such as market information, transactions, economic indicators and public opinions, and evaluating a relationship, a trend and an influence degree among the data through data arrangement, cleaning and exploratory analysis; real-time analysis is carried out according to the characteristics of the financial market, trend correlation between data and the influence degree on the market are comprehensively considered, the quality value of the data is evaluated, classification marking is carried out, and finally, the risk level possibly occurring in the financial market is predicted and warned in advance by analyzing the change of the quality value of the data within a period of time.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to the field of data analysis technology, and in particular to a financial risk prediction method and system based on big data. Background Art

[0002] Financial risk prediction based on big data refers to the process of identifying, evaluating and predicting various risks that may occur in the financial market by using large-scale data analysis technology and algorithms. First, big data technology allows financial institutions to collect and integrate large amounts of data from different sources, including market data, transaction data, economic indicators, public opinion data, etc. Secondly, by deeply mining and analyzing these data, potential connections and patterns between data can be discovered, thereby identifying potential financial risk factors. Finally, machine learning and statistical models are used to model and predict these factors, helping financial institutions to promptly discover and respond to possible risks, reduce financial market volatility, and ensure the stable operation of the financial system.

[0003] The prior art has the following deficiencies:

[0004] When using large-scale data analysis technology and algorithms to identify, evaluate and predict various risks that may occur in the financial market, as the amount of data in the financial market continues to increase, the quality value of the data will also become uneven. If the system cannot accurately identify data with higher quality value in a timely manner, it may lead to a reduction in the risk control of financial institutions. If the quality value of the data obtained by the system is low, the risk management system may not be able to identify and respond to potential risks in a timely manner, exposing financial institutions to unknown and uncontrollable risks, increasing the possibility of systemic risks and losses.

[0005] The quality value of data refers to the index obtained by comprehensively evaluating factors such as the strong correlation between data generated by the financial market and the degree of importance to business decisions. The higher the correlation between data, the greater the importance to business decisions, and the higher the quality value of data. In the financial field, the quality value of data is particularly important. If the quality value of data is high, financial institutions can assess risks more accurately and make better investment and business decisions, thereby reducing potential risks and losses. On the contrary, if the quality value of data is low, it may cause financial institutions to make wrong decisions, increasing the risk of systemic risks and losses. Therefore, for financial risk prediction and management, it is crucial to ensure the quality value of data. Summary of the invention

[0006] The purpose of the present invention is to provide a financial risk prediction method and system based on big data to address the deficiencies in the background technology.

[0007] In order to achieve the above-mentioned object, the present invention provides the following technical solution: a financial risk prediction method based on big data, comprising the following steps;

[0008] S1: Collect large-scale data related to the financial market, including market data, transaction data, economic indicators, and public opinion data, organize and clean the data, and improve the accuracy of the data;

[0009] S2: Conduct exploratory analysis on the collected data, visualize the relationships and trends between variables, and determine the degree of trend correlation between data;

[0010] S3: Based on the characteristics of the financial market, conduct real-time market analysis on the collected data and evaluate the impact of the data on the financial market;

[0011] S4: Comprehensively analyze the trend correlation between data and the impact of data on the financial market, and evaluate the value of data quality;

[0012] S5: Classify and label the data according to the evaluation results, and divide the data into high-quality data, medium-quality data, and low-quality data;

[0013] S6: Analyze the changes in the quality value of data collected over a period of time, and make early predictions and warnings on the risk levels in the financial market.

[0014] In a preferred embodiment, in S2, the trend correlation degree between the data is determined, and the data trend correlation index is obtained according to the trend correlation degree between the data. Then, the method for obtaining the data trend correlation index is:

[0015] Obtain two time series to be compared corresponding to the data acquisition time: X = (x1, x2, ..., xn) and Y = (y1, y2, ..., ym), where n and m represent the length of the series respectively;

[0016] Create a two-dimensional matrix D, where D[i][j] represents the distance between the i-th element of sequence X and the j-th element of sequence Y;

[0017] Calculate the dynamic regularization path, starting from D[1][1], and move along the minimum cost path of matrix D until reaching D[n][m];

[0018] Calculate the sum of the distances between all corresponding elements on the dynamic regularization path as the trend correlation index; calculate the distance matrix D[i][j]=d(xi,yj); where d(xi,yj) represents the distance between the i-th element xi of sequence X and the j-th element yj of sequence Y;

[0019] Dynamic programming is used to solve the minimum cost path. The path matrix is ​​initialized, P[1][1] = (1, 1). For P[i][j], the minimum value among D[i][j], D[i-1][j] and D[i][j-1] is selected, and the path matrix is ​​updated until P[n][m]. The trend correlation index is defined as the sum of the distances between all corresponding elements on the dynamic regularization path. Where Dt xy is the data trend correlation index, and p represents the dynamic regularization path.

[0020] In a preferred embodiment, in S3, the impact of data on the financial market is evaluated, including obtaining a data value composite index using the impact of data on the financial market. The method for obtaining the data value composite index is:

[0021] Mark the acquired data and use matrix X to represent their data matrix. Then the calculation expression of the correlation matrix R is: R ij =Corr(X i ,X j );where R ij Indicates the degree of correlation between data i and data j;

[0022] After building the data network, calculate the network degree centrality. The degree centrality of node i is The specific calculation expression is: In the formula, Represents the degree centrality of node i; the importance of the nodes is integrated to calculate the comprehensive importance of the nodes. The specific calculation expression is: In the formula, w k is the weight of the k-th centrality indicator, is the kth centrality index of node i, m is the number of centrality indexes, Indicates the comprehensive importance of node i; calculates the data value composite index, the specific calculation expression is: In the formula, yk m is the composite index of data value, N is the number of nodes in the network, Represents the comprehensive importance of node i.

[0023] In a preferred implementation, in S4, the trend correlation degree between data and the impact degree of data on the financial market are comprehensively analyzed to evaluate the data quality value, specifically:

[0024] The data trend correlation index and the data value composite index are normalized, and the quality value assessment coefficient of the large-scale data collected related to the financial market is calculated by using the normalized data trend correlation index and the data value composite index.

[0025] In a preferred embodiment, in S5, the data is classified and marked according to the evaluation result, and the data is divided into high-quality data, medium-quality data and low-quality data, specifically:

[0026] The obtained quality value assessment coefficient is compared with the gradient standard threshold, the gradient standard threshold includes a first standard threshold and a second standard threshold, and the first standard threshold is less than the second standard threshold, and the quality value assessment coefficient is compared with the first standard threshold and the second standard threshold respectively;

[0027] If the quality value assessment coefficient is greater than the second standard threshold, a data normal signal is issued and the corresponding data is marked as high-quality data;

[0028] If the quality value assessment coefficient is greater than the first standard threshold and less than or equal to the second standard threshold, no signal is sent and the corresponding data is marked as medium quality data;

[0029] If the quality value assessment coefficient is less than or equal to the first standard threshold, a data anomaly signal is issued and the corresponding data is marked as low-quality data.

[0030] In a preferred implementation, in S6, the changes in the data quality value collected over a period of time are analyzed, and the risk level in the financial market is predicted and warned in advance, specifically:

[0031] The data duration difference, water level smoothness value and overall performance anomaly index are normalized, and the accuracy coefficient of the geological data in the sub-monitoring area is calculated by the normalized data duration difference, water level smoothness value and overall performance anomaly index;

[0032] The accuracy coefficient of the geological data in the sub-monitoring area is compared with the accuracy threshold. If the accuracy coefficient of the geological data in the sub-monitoring area is greater than or equal to the accuracy threshold, a normal data signal is issued; if the accuracy coefficient of the geological data in the sub-monitoring area is less than the accuracy threshold, an abnormal data signal is issued.

[0033] In a preferred embodiment, in S6, the accuracy of the geological data in the sub-monitoring area within the specified time is analyzed, and the risk level of geological disasters in the sub-monitoring area is predicted in advance and an early warning is issued;

[0034] Set the financial market risk prediction interval, obtain the risk ratio of abnormal data signals within the financial market risk prediction interval, and analyze it, specifically:

[0035] If there is no data abnormality signal within the financial market risk prediction interval, no warning signal will be issued at this time;

[0036] If a data anomaly signal is issued within the financial market risk prediction interval, and the risk ratio of the data anomaly signal is less than the risk ratio threshold, a secondary warning signal is generated;

[0037] If a data anomaly signal is issued within the financial market risk prediction interval, and the risk ratio of the data anomaly signal is greater than or equal to the risk ratio threshold, a first-level warning signal is generated.

[0038] The present invention also provides a financial risk prediction system based on big data, including a data collection module, a data exploratory analysis module, a market real-time analysis module, a comprehensive analysis and evaluation module, a data classification and marking module, and a risk prediction and early warning module;

[0039] Data collection module: collects large-scale data related to the financial market, including market data, transaction data, economic indicators, and public opinion data, organizes and cleans the data, and improves the accuracy of the data;

[0040] Data exploratory analysis module: conduct exploratory analysis on the collected data, visualize the relationship and trend between various variables, and determine the degree of trend correlation between data;

[0041] Market real-time analysis module: Based on the characteristics of the financial market, the collected data is analyzed in real time to assess the impact of the data on the financial market;

[0042] Comprehensive analysis and evaluation module: conducts a comprehensive analysis of the trend correlation between data and the impact of data on the financial market, and evaluates the value of data quality;

[0043] Data classification and labeling module: classifies and labels the data according to the evaluation results, and divides the data into high-quality data, medium-quality data, and low-quality data;

[0044] Risk prediction and early warning module: Analyze the changes in the quality value of data collected over a period of time, and make early predictions and warnings on the risk levels in the financial market.

[0045] In the above technical solution, the technical effects and advantages provided by the present invention are:

[0046] 1. By collecting, cleaning and analyzing large-scale financial market-related data, the system can more accurately identify and assess various risks. Exploratory analysis and trend correlation assessment help identify patterns and trends in the data, while comprehensive analysis and classification marking steps further improve the quality value of the data, making risk prediction more accurate. The final early warning mechanism monitors changes in data quality value and promptly warns of possible risk events, helping financial institutions respond in a timely manner.

[0047] 2. Through comprehensive data collection and analysis, the present invention improves the financial institutions' ability to perceive market risks and reduces the possibility of systemic risks. At the same time, the prediction and early warning mechanism can give warnings before risks occur, allowing financial institutions to take timely countermeasures to reduce losses. The entire method has a clear process and rigorous logic, providing financial institutions with a set of comprehensive and effective risk management tools, which helps to improve the stability and sustainability of the financial market. BRIEF DESCRIPTION OF THE DRAWINGS

[0048] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the drawings required for use in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments recorded in the present invention. For ordinary technicians in this field, other drawings can also be obtained based on these drawings.

[0049] Figure 1 The figure is a flow chart of the method of the present invention.

[0050] Figure 2 It is a system module diagram of the present invention. DETAILED DESCRIPTION

[0051] In order to make the purpose, technical solution and advantages of the embodiments of the present invention clearer, the technical solution 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. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0052] Example 1

[0053] See also Figure 1 As shown, the financial risk prediction method based on big data described in this embodiment includes the following steps:

[0054] S1: Collect large-scale data related to the financial market, including market data, transaction data, economic indicators, and public opinion data, organize and clean the data, and improve the accuracy of the data;

[0055] S2: Conduct exploratory analysis on the collected data, visualize the relationships and trends between variables, and determine the degree of trend correlation between data;

[0056] S3: Based on the characteristics of the financial market, conduct real-time market analysis on the collected data and evaluate the impact of the data on the financial market;

[0057] S4: Comprehensively analyze the degree of trend correlation between data and the degree of impact of data on the financial market, and evaluate the data quality value;

[0058] S5: According to the evaluation results, classify and label the data, and divide the data into high-quality data, medium-quality data, and low-quality data;

[0059] S6: Analyze the change situation of the data quality value collected within a period of time, and predict and give early warnings of the risk levels emerging in the financial market.

[0060] Among them, in S1, collect a large amount of data related to the financial market, including market quotation data, transaction data, economic indicators, and public opinion data, and sort out and clean the data to improve the accuracy of the data;

[0061] By sorting out and cleaning the data to improve the accuracy of the data, the problem of inaccurate or incomplete data can be solved to a certain extent. The cleaned data helps to reduce the possibility that the risk management system cannot identify potential risks.

[0062] First, collect text data related to the financial market, such as news reports, social media comments, analyst reports, etc. These data may contain descriptions and views on aspects such as market sentiment, events, and company performance.

[0063] Clean the collected text data, including removing HTML tags, special characters, stop words (such as "of", "is", etc.), and performing preprocessing steps such as stemming or lemmatization to reduce noise and improve data quality.

[0064] Extract features from the cleaned text data through text mining techniques. Commonly used features include the bag of words model, TF-IDF (Term Frequency-Inverse Document Frequency) weights, etc. These features can help machine learning models understand the text content and conduct analysis.

[0065] Use sentiment analysis techniques to identify the sentiment polarity of the text data, that is, judge whether the sentiment expressed by the text is positive, negative, or neutral. This can help identify market sentiment and thus affect the trend of the financial market.

[0066] Conduct entity recognition to identify specific entities mentioned in the text, such as companies, people, locations, etc., to help understand the subjects and relationships involved in the text.

[0067] Use topic modeling techniques (such as Latent Dirichlet Allocation, LDA) to extract and classify text topics, thereby identifying hidden themes and topics in text data.

[0068] The extracted text features are combined with other data (such as market data, economic indicators, etc.) to build a prediction model to predict the trend of the financial market or the impact of specific events.

[0069] In S2, exploratory analysis is performed on the collected data to visualize the relationships and trends among the variables and determine the degree of trend correlation among the data.

[0070] Understand the basic information of the data set, including the meaning of the fields, data types, etc. Perform preliminary cleaning and preprocessing of the data to handle missing values, outliers, etc.

[0071] Choose appropriate visualization tools according to data type and analysis purpose, such as Matplotlib, Seaborn, Plotly, etc. Choose appropriate chart types according to variable types and relationships, such as scatter plots, line charts, box plots, etc.

[0072] For continuous variables, use scatter plots or correlation matrices to explore the linear correlation between variables. Scatter plots can intuitively show the relationship between two variables, while correlation matrices can show the correlation between multiple variables at once.

[0073] For the relationship between categorical variables and continuous variables, box plots or violin plots can be used to show the data distribution at different classification levels.

[0074] If the data has time series characteristics, you can use a line chart or area chart to analyze the trend of the variable over time. This helps to discover the seasonality, trend and other characteristics of the data.

[0075] If the data does not have time series characteristics, you can analyze the changing trend of the variable by sorting the data in a certain order and drawing a line graph.

[0076] For multidimensional data, you can use heat maps or cluster maps to explore possible patterns in the data. Heat maps can intuitively show the correlation between different variables, while cluster maps can help discover clustering structures in the data.

[0077] In the process of visual analysis, key information and insights are continuously extracted to understand the patterns and characteristics in the data. This information can help us deeply understand the data and provide important reference for subsequent analysis and prediction.

[0078] Interpret and summarize the results of visual analysis, write reports or create visual charts to share the analysis results and findings with the team or stakeholders.

[0079] Determine the trend correlation degree between data, and obtain the data trend correlation index through the trend correlation degree between data. The method for obtaining the data trend correlation index is:

[0080] Obtain two time series to be compared corresponding to the data acquisition time: X = (x1, x2, ..., xn) and Y = (y1, y2, ..., ym), where n and m represent the lengths of the sequences respectively.

[0081] Create a two-dimensional matrix D where D[i][j] represents the distance between the i-th element of sequence X and the j-th element of sequence Y.

[0082] Calculate the dynamic regularization path, starting from D[1][1], and move along the minimum cost path of matrix D until reaching D[n][m]. Define a distance metric function, such as Euclidean distance, Manhattan distance, etc., to calculate the distance between each pair of elements in two sequences.

[0083] Calculate the sum of the distances between all corresponding elements on the dynamic regularization path as the trend correlation index. Calculate the distance matrix D[i][j]=d(xi,yj); where d(xi,yj) represents the distance between the i-th element xi of sequence X and the j-th element yj of sequence Y.

[0084] Dynamic programming solves the minimum cost path and initializes the path matrix, P[1][1] = (1,1). For P[i][j], select the minimum value among D[i][j], D[i-1][j] and D[i][j-1], and update the path matrix until P[n][m]. Define the trend correlation index as the sum of the distances between all corresponding elements on the dynamic regularization path. Where Dt xy is the data trend correlation index, and p represents the dynamic regularization path.

[0085] The larger the data trend correlation index is, the more similar or matching the trends between the data are. In financial risk prediction, this may mean that the trends between different data sources are more consistent, which helps to improve the accuracy and reliability of the prediction model. Specifically:

[0086] When the data trend correlation index is large, it means that the fluctuations and change trends between different data are similar, which may reflect the consistent behavior in the financial market. Therefore, the data quality may be higher and the information provided is more consistent and reliable.

[0087] The higher the trend correlation between data, the higher the accuracy of the forecasting model will generally be, because the consistency of data can reduce the noise in the model, making the model more capable of capturing the real market dynamics and risk trends.

[0088] High trend correlation between data means high consistency and reliability of information, which can improve the reliability of risk assessment and prediction. Financial institutions can rely on this data for risk management decisions with greater confidence.

[0089] S3: Based on the characteristics of the financial market, conduct real-time market analysis on the collected data and evaluate the impact of the data on the financial market;

[0090] Assess the impact of data on the financial market, including obtaining a composite index of data value using the impact of data on the financial market. The method for obtaining the composite index of data value is as follows:

[0091] Mark the acquired data and use matrix X to represent their data matrix. Then the calculation expression of the correlation matrix R is: R ij =Corr(X i ,X j );where R ij It indicates the degree of correlation between data i and data j, which is usually measured by correlation coefficient or mutual information. After building the data network, the degree centrality of the network is calculated. The degree centrality of node i is It can be obtained by calculating the degree of node i. The specific calculation expression is: In the formula, It represents the degree centrality of node i, that is, the sum of the number of connections between node i and other nodes. It reflects the degree of direct connection of the node in the network.

[0092] The importance of nodes can be combined and weighted summed according to needs, different weights can be given to different importance indicators to calculate the comprehensive importance of nodes. In the formula, w k is the weight of the k-th centrality indicator, is the kth centrality index of node i, and m is the number of centrality indexes; the data value composite index is calculated, and the specific calculation expression is: In the formula, yk m is the composite index of data value, N is the number of nodes in the network, It represents the comprehensive importance of node i, which is the result of combining various centrality indicators of the node. It can reflect the comprehensive influence and importance of the node in the entire network.

[0093] The larger the data value composite index is, the higher the quality of the data in the financial market is. High-value data is usually more comprehensive and accurate, and can provide more useful information to help financial institutions better understand market conditions and risk situations.

[0094] High-value data is more strongly correlated with other data in the financial market, and can better reflect the potential correlations and changing trends in the market, providing a more reliable basis for risk prediction.

[0095] High-value data is usually more stable and less susceptible to external interference or abnormal situations. It can provide more reliable data support and reduce the risks of financial institutions in the decision-making process.

[0096] High-value data can usually reflect market changes in a timely manner, has high timeliness, and can help financial institutions make decisions and respond to market risks more quickly.

[0097] S4: Comprehensively analyze the trend correlation between data and the impact of data on the financial market, and evaluate the value of data quality.

[0098] The data trend correlation index and the data value composite index are normalized, and the quality value assessment coefficient of the large-scale data collected related to the financial market is calculated by using the normalized data trend correlation index and the data value composite index.

[0099] For example, the present invention can use the following formula to calculate the quality value evaluation coefficient of the collected large-scale data related to the financial market, and the calculation expression is: In the formula, nh m is the quality value assessment coefficient, Dt xy is the data trend correlation index, yk m is the composite index of data value, a 1 、a 2 is the ratio coefficient of the data trend correlation index and the data value composite index, and a 2 >a 1 >0;

[0100] The data trend correlation index and the data value composite index are both positively correlated with the quality value assessment coefficient, that is, as the data trend correlation index and the data value composite index increase, the quality value assessment coefficient also increases accordingly, and the quality value of the large-scale data related to the collected financial market is higher.

[0101] S5: Based on the evaluation results, classify and label the data into high-quality data, medium-quality data, and low-quality data. Specifically:

[0102] Compare the obtained quality value assessment coefficient with the gradient standard threshold, the gradient standard threshold in the present application includes a first standard threshold and a second standard threshold, and the first standard threshold is less than the second standard threshold, and compare the quality value assessment coefficient with the first standard threshold and the second standard threshold respectively;

[0103] If the quality value assessment coefficient is greater than the second standard threshold, it means that the quality value of the large-scale data collected related to the financial market is higher. At this time, a data normal signal is issued and the corresponding data is marked as high-quality data. Relevant personnel can directly use this data for financial risk prediction and analysis without additional processing;

[0104] If the quality value assessment coefficient is greater than the first standard threshold and less than or equal to the second standard threshold, it means that the quality value of the large-scale data collected related to the financial market is average. In this case, no signal is issued and the corresponding data is marked as medium-quality data. Relevant personnel can further review the data to ensure its accuracy and reliability, and then use it for financial risk analysis as needed;

[0105] If the quality value assessment coefficient is less than or equal to the first standard threshold, it means that the quality value of the large-scale data collected related to the financial market is lower. At this time, a data anomaly signal is issued and the corresponding data is marked as low-quality data. Relevant personnel need to further investigate, review and cleanse the data to eliminate data errors or anomalies in order to improve the quality of the data.

[0106] In this embodiment, large-scale data related to the financial market is collected, sorted and cleaned to improve the accuracy of the data; exploratory analysis is performed on the collected data, the relationship and trend between each variable are visualized, and the degree of trend correlation between the data is determined; according to the characteristics of the financial market, the market value of the collected data is analyzed to evaluate the impact of the data on the financial market; the degree of trend correlation between the data and the impact of the data on the financial market are comprehensively analyzed to evaluate the data quality value; according to the evaluation results, the data is classified and labeled to divide the data into high-quality data, medium-quality data and low-quality data, which can effectively improve the quality value of financial market data and ensure that the data can be accurately used for financial risk prediction and analysis.

[0107] Example 2

[0108] S6: Analyze the changes in the quality value of data collected over a period of time, and make early predictions and warnings on the risk levels in the financial market, specifically:

[0109] When the quality value of data related to the financial market collected over a period of time has been at a low level, it means that the possibility of risk events in the financial market is higher, specifically:

[0110] Set a financial market risk prediction interval. The time length corresponding to the financial market risk prediction interval is fixed, and its range changes over time, always ending at the real-time point. Ensure real-time monitoring and assessment of financial market risks, timely identify potential risks and take necessary countermeasures to mitigate possible losses and impacts, mainly including:

[0111] Obtain the risk ratio of abnormal data signals within the risk prediction interval of the financial market and conduct analysis, specifically:

[0112] If there is no abnormal data signal within the financial market risk prediction interval, it means that the possibility of risk events in the financial market is at a low level. At this time, no early warning signal will be issued, and the market situation can continue to be monitored and observed, but no special response measures are required;

[0113] If a data anomaly signal is issued within the risk prediction interval of the financial market, and the risk ratio of the data anomaly signal is less than the risk ratio threshold, it means that the possibility of a risk event occurring in the financial market is at a medium level. At this time, a secondary warning signal is generated to remind market participants that there may be potential risks, but the risks have not yet reached a high level and are still within a controllable range. It is recommended to closely monitor market conditions, invest or trade prudently, and adopt appropriate risk management strategies to prevent possible risks.

[0114] If an abnormal data signal is issued within the risk prediction interval of the financial market, and the risk ratio of the abnormal data signal is greater than or equal to the risk ratio threshold, it means that the possibility of a risk event occurring in the financial market is at a high level. At this time, a first-level warning signal is generated, indicating that there may be major risk events in the financial market, which requires market participants to be highly vigilant. It is recommended to take timely and effective risk control measures, which may include reducing positions, closing positions or taking hedging measures to reduce the impact of risks on investment portfolios or trading strategies.

[0115] The risk ratio threshold is set by personnel in this field based on the specific circumstances of changes in financial market data and will not be elaborated here.

[0116] Specifically, the risk level of the first-level warning signal is higher than that of the second-level warning signal. The first-level warning signal indicates that the financial market may face major risk events, which requires market participants to be highly vigilant. The second-level warning signal indicates that there are potential risks in the market, but it has not yet reached the level of causing major impact. At this level, the risk is still within the controllable range, but cautious investment and trading strategies need to be adopted.

[0117] In this embodiment, according to the data quality and value changes of the financial market, the financial market risk prediction interval is set. Within the financial market risk prediction interval, the risk ratio of the data abnormality signal is obtained. If no data abnormality signal is issued, it means that the possibility of a risk event in the financial market is low, and no warning signal is issued, and the market situation continues to be monitored. If a data abnormality signal is issued, but the risk ratio is less than the preset threshold, a secondary warning signal is generated to remind market participants that there is a potential risk, but it is still within a controllable range, and it is recommended to adopt a cautious investment or trading strategy. If the risk ratio of the data abnormality signal is greater than or equal to the risk ratio threshold, a primary warning signal is generated, indicating that there may be a major risk event, which requires market participants to be highly vigilant, and it is recommended to take timely and effective risk control measures to reduce possible losses. At the same time, the risk level of the primary warning signal is higher than the risk level of the secondary warning signal. Through the early warning system, market participants can better understand market risks and make corresponding response preparations, thereby reducing market uncertainty and improving investor confidence and market stability.

[0118] Example 3

[0119] See also Figure 2 As shown, the financial risk prediction system based on big data described in this embodiment includes a data collection module, a data exploratory analysis module, a market real-time analysis module, a comprehensive analysis and evaluation module, a data classification and marking module, and a risk prediction and early warning module;

[0120] Data collection module: collects large-scale data related to the financial market, including market data, transaction data, economic indicators, and public opinion data, organizes and cleans the data, and improves the accuracy of the data;

[0121] Data exploratory analysis module: conduct exploratory analysis on the collected data, visualize the relationship and trend between various variables, and determine the degree of trend correlation between data;

[0122] Market real-time analysis module: Based on the characteristics of the financial market, the collected data is analyzed in real time to assess the impact of the data on the financial market;

[0123] Comprehensive analysis and evaluation module: conducts a comprehensive analysis of the trend correlation between data and the impact of data on the financial market, and evaluates the value of data quality;

[0124] Data classification and labeling module: classifies and labels the data according to the evaluation results, and divides the data into high-quality data, medium-quality data, and low-quality data;

[0125] Risk prediction and early warning module: Analyze the changes in the quality value of data collected over a period of time, and make early predictions and warnings on the risk levels in the financial market.

[0126] The above formulas are all dimensionless and numerical calculations. The formula is a formula for the most recent real situation obtained by collecting a large amount of data and performing software simulation. The preset parameters in the formula are set by technicians in this field according to actual conditions.

[0127] The above embodiments can be implemented in whole or in part by software, hardware, firmware or any other combination. When implemented by software, the above embodiments can be implemented in whole or in part in the form of a computer program product. The computer program product includes one or more computer instructions or computer programs. When the computer instructions or computer programs are loaded or executed on a computer, the process or function described in the embodiment of the present application is generated in whole or in part. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions can be stored in a computer-readable storage medium, or transmitted from one computer-readable storage medium to another computer-readable storage medium. For example, the computer instructions can be transmitted from one website site, computer, server or data center to another website site, computer, server or data center by wired or wireless (e.g., infrared, wireless, microwave, etc.). The computer-readable storage medium can be any available medium that a computer can access or a data storage device such as a server or data center that contains one or more available media sets. The available medium can be a magnetic medium (e.g., a floppy disk, a hard disk, a tape), an optical medium (e.g., a DVD), or a semiconductor medium. The semiconductor medium can be a solid-state hard disk.

[0128] It should be understood that the term "and / or" in this article is only a description of the association relationship of associated objects, indicating that there can be three relationships. For example, A and / or B can represent: A exists alone, A and B exist at the same time, and B exists alone. A and B can be singular or plural. In addition, the character " / " in this article generally indicates that the associated objects before and after are in an "or" relationship, but it may also indicate an "and / or" relationship. Please refer to the context for specific understanding.

[0129] It should be understood that in the various embodiments of the present application, the size of the serial numbers of the above-mentioned processes does not mean the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of the present application.

[0130] Those of ordinary skill in the art will appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of this application.

[0131] The above is only a specific implementation of the present application, but the protection scope of the present application is not limited thereto. Any person skilled in the art who is familiar with the present technical field can easily think of changes or substitutions within the technical scope disclosed in the present application, which should be included in the protection scope of the present application. Therefore, the protection scope of the present application should be based on the protection scope of the claims.

Claims

1. A financial risk prediction method based on big data, characterized by: The steps include: S1: Collect large-scale data related to the financial market, including market data, transaction data, economic indicators, and public opinion data, organize and clean the data, and improve the accuracy of the data; S2: Conduct exploratory analysis on the collected data, visualize the relationships and trends between variables, and determine the degree of trend correlation between data; S3: Based on the characteristics of the financial market, conduct real-time market analysis on the collected data and evaluate the impact of the data on the financial market; S4: Comprehensively analyze the trend correlation between data and the impact of data on the financial market, and evaluate the value of data quality; S5: Classify and label the data according to the evaluation results, and divide the data into high-quality data, medium-quality data, and low-quality data; S6: Analyze the changes in the quality value of data collected over a period of time, and make early predictions and warnings on the risk levels in the financial market.

2. The financial risk prediction method based on big data according to claim 1 is characterized by: In S2, the trend correlation degree between the data is determined, and the data trend correlation index is obtained according to the trend correlation degree between the data. Then, the method for obtaining the data trend correlation index is: Obtain two time series to be compared corresponding to the data acquisition time: X = (x1, x2, ..., xn) and Y = (y1, y2, ..., ym), where n and m represent the length of the series respectively; Create a two-dimensional matrix D, where D[i][j] represents the distance between the i-th element of sequence X and the j-th element of sequence Y; Calculate the dynamic regularization path, starting from D[1][1], and move along the minimum cost path of matrix D until reaching D[n][m]; Calculate the sum of the distances between all corresponding elements on the dynamic regularization path as the trend correlation index; calculate the distance matrix D[i][j]=d(xi,yj); where d(xi,yj) represents the distance between the i-th element xi of sequence X and the j-th element yj of sequence Y; Dynamic programming is used to solve the minimum cost path. The path matrix is ​​initialized, P[1][1] = (1, 1). For P[i][j], the minimum value among D[i][j], D[i-1][j] and D[i][j-1] is selected, and the path matrix is ​​updated until P[n][m]. The trend correlation index is defined as the sum of the distances between all corresponding elements on the dynamic regularization path. Where Dt xy is the data trend correlation index, and p represents the dynamic regularization path.

3. The financial risk prediction method based on big data according to claim 2 is characterized in that: In S3, the impact of data on the financial market is evaluated, including obtaining a composite index of data value using the impact of data on the financial market. The method for obtaining the composite index of data value is as follows: Mark the acquired data and use matrix X to represent their data matrix. Then the calculation expression of the correlation matrix R is: R ij =CorrX i ,X j ; In the formula, R ij Indicates the degree of correlation between data i and data j; After building the data network, calculate the network degree centrality. The degree centrality of node i is The specific calculation expression is: In the formula, Represents the degree centrality of node i; the importance of the nodes is integrated to calculate the comprehensive importance of the nodes. The specific calculation expression is: In the formula, w k is the weight of the k-th centrality indicator, is the kth centrality index of node i, m is the number of centrality indexes, Indicates the comprehensive importance of node i; calculates the data value composite index, the specific calculation expression is: In the formula, yk m is the composite index of data value, N is the number of nodes in the network, Represents the comprehensive importance of node i.

4. The financial risk prediction method based on big data according to claim 3 is characterized by: In S4, the trend correlation between data and the impact of data on the financial market are comprehensively analyzed to evaluate the value of data quality, specifically: The data trend correlation index and the data value composite index are normalized, and the quality value assessment coefficient of the large-scale data collected related to the financial market is calculated by using the normalized data trend correlation index and the data value composite index.

5. The financial risk prediction method based on big data according to claim 4 is characterized in that: In S5, according to the evaluation results, the data is classified and marked, and the data is divided into high-quality data, medium-quality data, and low-quality data, specifically: The obtained quality value assessment coefficient is compared with the gradient standard threshold, the gradient standard threshold includes a first standard threshold and a second standard threshold, and the first standard threshold is less than the second standard threshold, and the quality value assessment coefficient is compared with the first standard threshold and the second standard threshold respectively; If the quality value assessment coefficient is greater than the second standard threshold, a data normal signal is issued and the corresponding data is marked as high-quality data; If the quality value assessment coefficient is greater than the first standard threshold and less than or equal to the second standard threshold, no signal is sent and the corresponding data is marked as medium quality data; If the quality value assessment coefficient is less than or equal to the first standard threshold, a data anomaly signal is issued and the corresponding data is marked as low-quality data.

6. The financial risk prediction method based on big data according to claim 5 is characterized by: In S6, the changes in the quality value of the data collected over a period of time are analyzed, and the risk level in the financial market is predicted and warned in advance, specifically: The data duration difference, water level smoothness value and overall performance anomaly index are normalized, and the accuracy coefficient of the geological data in the sub-monitoring area is calculated by the normalized data duration difference, water level smoothness value and overall performance anomaly index; The accuracy coefficient of the geological data in the sub-monitoring area is compared with the accuracy threshold. If the accuracy coefficient of the geological data in the sub-monitoring area is greater than or equal to the accuracy threshold, a normal data signal is issued; if the accuracy coefficient of the geological data in the sub-monitoring area is less than the accuracy threshold, an abnormal data signal is issued.

7. The financial risk prediction method based on big data according to claim 6 is characterized by: In S6, the accuracy of geological data in the sub-monitoring area within the specified time is analyzed, and the risk level of geological disasters in the sub-monitoring area is predicted in advance and an early warning is issued; Set the financial market risk prediction interval, obtain the risk ratio of abnormal data signals within the financial market risk prediction interval, and analyze it, specifically: If there is no data abnormality signal within the financial market risk prediction interval, no warning signal will be issued at this time; If a data anomaly signal is issued within the financial market risk prediction interval, and the risk ratio of the data anomaly signal is less than the risk ratio threshold, a secondary warning signal is generated; If a data anomaly signal is issued within the financial market risk prediction interval, and the risk ratio of the data anomaly signal is greater than or equal to the risk ratio threshold, a first-level warning signal is generated.

8. A financial risk prediction system based on big data, used to implement the financial risk prediction method based on big data according to any one of claims 1 to 7, characterized in that: It includes data collection module, data exploratory analysis module, market real-time analysis module, comprehensive analysis and evaluation module, data classification and marking module and risk prediction and early warning module; Data collection module: collects large-scale data related to the financial market, including market data, transaction data, economic indicators, and public opinion data, organizes and cleans the data, and improves the accuracy of the data; Data exploratory analysis module: conduct exploratory analysis on the collected data, visualize the relationship and trend between various variables, and determine the degree of trend correlation between data; Market real-time analysis module: Based on the characteristics of the financial market, the collected data is analyzed in real time to assess the impact of the data on the financial market; Comprehensive analysis and evaluation module: conducts a comprehensive analysis of the trend correlation between data and the impact of data on the financial market, and evaluates the value of data quality; Data classification and labeling module: classifies and labels the data according to the evaluation results, and divides the data into high-quality data, medium-quality data, and low-quality data; Risk prediction and early warning module: Analyze the changes in the quality value of data collected over a period of time, and make early predictions and warnings on the risk levels in the financial market.

Citation Information

Cited By

  • Financial risk assessment method based on big data

    CN120823045A