A Method and System for Constructing a Model for Predicting Enterprise Financial Data

Abstract

The present invention provides a method and system for constructing a model for predicting enterprise financial data, which relates to the technical field of data science. The present invention extracts the financial data of the enterprise in the previous year's quarters, calculates the expense-income ratio, gross profit margin, profit conversion rate, and free cash flow ratio of the enterprise, and then generates a financial health index. Combining with the historical financial data of other enterprises in the same industry, calculates the revenue change rate and average profit margin of the industry, generates a financial correction factor, uses a neural network model to predict the financial security index, and corrects it through an adjustment factor to obtain the final corrected value of the financial security index. Sets a financial risk threshold to judge the financial health status of the enterprise. The present invention constructs a financial prediction model by collecting and processing enterprise financial data and combining with industry financial correction factors, realizing the dynamic evaluation and risk warning of the enterprise's financial health status.

Description

Technical Field

[0001] The present invention relates to the technical field of data science, and particularly to a method and system for constructing a model for predicting enterprise financial data. Background Art

[0002] In today's complex and ever-changing business environment, accurately predicting a company's financial data is of crucial significance to management, investors, and other stakeholders. Financial data prediction can not only help enterprises cope with uncertainties but also provide a reliable basis for strategic decision-making. The market environment is unpredictable, and external factors such as the macroeconomic situation, industry policies, and competitive landscape will all affect a company's financial performance. Through prediction, enterprises can identify potential financial risks and opportunities in advance and take corresponding measures to address them. For example, predicting fluctuations in cash flow helps avoid shortages of funds and formulate financing plans in advance. Predicting financial data such as a company's revenue, expenses, and profits can provide guidance for resource allocation. Enterprises can prepare budgets based on the predicted results to ensure the rational use of funds and resources and maximize capital returns.

[0003] In existing financial analysis, it often relies on simple financial indicators and linear models, which often fail to fully capture the non-linear relationships between a company's financial data and are difficult to comprehensively reflect the complexity and dynamic changes of a company's financial situation. Moreover, traditional financial analysis methods mostly rely on historical data and periodic reports, such as quarterly statements and annual reports. Therefore, their analysis results are often lagging and cannot reflect a company's financial health in real time. Enterprises may not be able to detect and respond to potential risks in a timely manner before financial problems occur, resulting in delayed decision-making and affecting the flexibility and adaptability of enterprises.

[0004] Traditional financial analysis methods often focus on past financial data and lack the ability of forward-looking prediction. With the development of technology, enterprises need models that can predict future financial conditions, cash flows, profitability, and other indicators to help enterprises make more accurate decisions. However, existing financial analysis methods usually lack sufficient predictive ability and cannot effectively support long-term strategic decision-making.

[0005] Therefore, it is necessary to provide a method and system for constructing a model for predicting enterprise financial data to solve the above problems.

[0006] The above information disclosed in the background art section is only used to enhance the understanding of the background of the present disclosure. Therefore, it may include information that does not constitute the prior art known to those of ordinary skill in the art. Summary of the Invention

[0007] The purpose of the present invention is to provide a method and system for coal mine supervision based on video images to solve the problems raised in the above background art.

[0008] To achieve the above object, the present invention provides the following technical solutions:

[0009] A method for constructing a model for predicting enterprise financial data, the specific steps including:

[0010] Step 1: Based on the enterprise financial data of the previous year's quarters of the enterprise to be predicted, extract characteristic parameter data, and the characteristic parameter data includes the operating income, total expenditure, cost of sales, total profit, net profit, total liabilities and free cash flow of the enterprise;

[0011] Step 2: Preprocess the collected characteristic parameter data, and combine the operating income, total expenditure, cost of sales, total profit, net profit, total liabilities and free cash flow of the enterprise after preprocessing, and perform permutation and combination to generate the expense input ratio, gross profit margin, profit conversion rate and free cash flow ratio of the enterprise finance respectively. The preprocessing includes data cleaning and normalization of the collected characteristic parameter data;

[0012] Step 3: Use the generated expense input ratio, gross profit margin, profit conversion rate and free cash flow ratio of the enterprise finance to calculate the financial health index of the enterprise;

[0013] Step 4: Obtain the historical financial data of other enterprise companies in the industry to which the enterprise to be predicted belongs, calculate the revenue change rate and average profit rate of the industry to which the enterprise to be predicted belongs, and combine the revenue change rate and average profit rate to generate the financial correction factor of the industry to which the enterprise to be predicted belongs;

[0014] Step 5: Establish a financial security prediction model, input the characteristic parameter data of the enterprise to be predicted into the financial security prediction model to obtain the financial security index, and use the calculated financial correction factor and financial health index to correct the financial security index to obtain the corrected value of the financial security index;

[0015] Step 6: Set a financial risk threshold, compare the corrected value of the financial security index with the financial risk threshold, and judge the health and security status of the enterprise financial data.

[0016] Further, the methods for data cleaning and normalization of the collected characteristic parameter data are as follows:

[0017] Data cleaning of the characteristic parameter data includes the detection and deletion of outliers and duplicate data, and the processing of missing values. Specifically, statistical methods are used to identify outliers and duplicate data in the characteristic parameter data, delete the outliers and duplicate data in the characteristic parameter data, and use the mean, median or mode of the characteristic parameter data to fill the missing values in the characteristic parameter data;

[0018] The characteristic parameter data is standardized by min-max normalization, scaling the data to the range [0, 1] so that all characteristic parameter data has the same scale. The formula is as follows:

[0019]

[0020] where X′ is the characteristic parameter data after normalization, X is the original characteristic parameter data, X min is the minimum value of the same type of characteristic parameter data in the dataset, X max is the maximum value of the same type of characteristic parameter data in the dataset.

[0021] Furthermore, permutations and combinations are performed to generate the expense input ratio, gross profit margin, profit conversion rate, and free cash flow ratio of the enterprise's finance. The formula is as follows:

[0022]

[0023]

[0024] The expense-income ratio is used to measure the enterprise's cost management and expense control. The gross profit margin is used to measure the enterprise's core profitability. The profit conversion rate is used to measure the conversion efficiency of the enterprise from total profit to net profit. The free cash flow ratio is used to measure the enterprise's debt repayment ability and financial risk.

[0025] Furthermore, calculate the financial health index of the enterprise. The method is as follows:

[0026] Define the expense-income ratio, gross profit margin, profit conversion rate, and free cash flow ratio of the enterprise's finance as letters a, b, c, and d respectively, and comprehensively calculate the financial health index of the enterprise. The formula is as follows:

[0027]

[0028] where JK represents the financial health index of the enterprise, and a, b, c, and d represent the expense-income ratio, gross profit margin, profit conversion rate, and free cash flow ratio of the enterprise's finance respectively.

[0029] Furthermore, combine the revenue change rate and the average profit rate to generate the financial correction factor for the industry to which the enterprise to be predicted belongs. The method is as follows:

[0030] Obtain the annual business revenues of N other enterprise companies in the industry last year and the year before last, calculate the absolute value of the difference between the annual business revenue last year and the year before last and the ratio to the annual business revenue last year, and take the average value as the revenue change rate. The formula is as follows:

[0031]

[0032] Among them, b represents the revenue change rate of the industry, are the operating revenues of the i-th company of other enterprises in the last year and the previous year respectively, i represents the index of other enterprise companies in the industry, i ∈ [1, N], and N is the number of other enterprise companies in the industry;

[0033] Obtain the net profits and operating revenues of N other enterprise companies in the industry in the last year, calculate their ratios, and take the average value as the average profit rate. The formula is:

[0034]

[0035] Among them, a represents the average profit rate of the industry, is the net profit of the i-th company of other enterprises in the last year;

[0036] Calculate the financial correction factor based on the collected revenue change rate and the average profit rate of the industry. The formula is:

[0037]

[0038] Among them, β represents the financial correction factor.

[0039] Furthermore, establish a financial security prediction model to obtain the corrected value of the financial security index of the enterprise. The method is:

[0040] Establish a financial security prediction model, with the characteristic parameter data of the enterprise as the input and the financial security index of the enterprise as the output. The financial security prediction model adopts a neural network convolutional structure, including an input layer, a hidden layer, and an output layer. The input layer is responsible for receiving the processed characteristic parameter data; the hidden layer is used to process the characteristic parameter data. By applying multiple convolutional kernels, the hidden layer can identify complex characteristic patterns. By using the ReLu activation function, introducing a non-linear relationship enables the model to fit complex characteristic relationships; an independent neuron is set in the output layer, which is responsible for converting the local and high-level characteristic representations extracted by the hidden layer into the final prediction result;

[0041] Then, use the calculated financial correction factor and financial health index to correct the financial security index to obtain the corrected value of the financial security index. The formula for calculating the corrected value of the financial security index is:

[0042] R safe ′ = e Jk+β *R safe

[0043] Among them, JK represents the financial health index, R safe ′ is the corrected value of the financial security index, R safeis the financial security index.

[0044] Further, compare the corrected value of the financial security index output by the model with the financial risk threshold to judge the health status of the enterprise's financial data. The logical formula is:

[0045]

[0046] where Q represents the logical value for judging the health status of the enterprise's financial data. When Q = 0, the corrected value of the enterprise's financial security index does not exceed the financial risk threshold, indicating that there is no potential risk in the enterprise's financial data at this time and it can continue to be maintained; when Q = 1, the corrected value of the enterprise's financial security index exceeds the financial risk threshold, indicating that there is a risk in the enterprise's financial data at this time and it needs to be monitored and adjusted; G health is the financial risk threshold.

[0047] The present invention also provides a system for constructing a model for predicting enterprise financial data. The construction system is used to execute the above-mentioned method for constructing a model for predicting enterprise financial data, including:

[0048] An enterprise financial feature data acquisition module, which is used to extract feature parameter data based on the enterprise's financial data in the previous year's quarters of the enterprise to be predicted. The feature parameter data includes the enterprise's operating income, total expenditure, cost of sales, total profit, net profit, total liabilities, and free cash flow;

[0049] A feature data preprocessing module, which is used to preprocess the collected feature parameter data. Combining the enterprise's operating income, total expenditure, cost of sales, total profit, net profit, total liabilities, and free cash flow after preprocessing, respectively generate the expense input ratio, gross profit margin, profit conversion rate, and free cash flow ratio of the enterprise's finance. The preprocessing includes data cleaning and normalization of the collected feature parameter data;

[0050] A financial health index generation module, which is used to calculate the financial health index of the enterprise by using the generated expense input ratio, gross profit margin, profit conversion rate, and free cash flow ratio of the enterprise's finance;

[0051] A financial correction factor generation module, which is used to obtain the historical financial data of other enterprise companies in the industry to which the enterprise to be predicted belongs, calculate the revenue change rate and average profit rate of the industry to which the enterprise to be predicted belongs, and generate the financial correction factor of the industry to which the enterprise to be predicted belongs in combination with the revenue change rate and average profit rate;

[0052] Construct a financial security prediction model module, which is used to establish a financial security prediction model. Input the characteristic parameter data of the enterprise to be predicted into the financial security prediction model to obtain a financial security index, and use the calculated financial correction factor and financial health index to correct the financial security index to obtain a corrected value of the financial security index;

[0053] A risk warning module, which is used to set a financial risk threshold, compare the corrected value of the financial security index with the financial risk threshold, and judge the health and security status of the enterprise's financial data.

[0054] Compared with the prior art, the beneficial effects of the present invention are as follows:

[0055] Through a series of data preprocessing and calculation methods, combined with the financial data of the industry, the present invention accurately evaluates and predicts the financial health status of the enterprise. Through feature extraction and preprocessing of the enterprise's financial data, key financial indicators such as expense input ratio, gross profit margin, profit conversion rate, and free cash flow ratio are generated, and further the financial health index of the enterprise is calculated. Combining the industry's revenue change rate and average profit rate to generate a financial correction factor, and correcting the financial security index through a neural network model, ultimately helping the enterprise accurately identify its financial security risks, and being able to provide a more accurate financial health assessment and risk warning for the enterprise. Through the quantified financial health index and the corrected value of the security index, the financial status of the enterprise can be dynamically monitored and adjusted according to industry standards to ensure timely discovery of potential financial problems, thereby helping the enterprise optimize financial decisions, reduce financial risks, and enhance its long-term development stability;

[0056] In addition, the present invention also calculates the financial correction factor by introducing the historical financial data of the industry, and corrects the financial security index according to this factor and the calculated financial health index, so that the judgment standard of the enterprise's financial health can be dynamically adjusted according to the industry characteristics and economic environment changes. This correction mechanism makes the judgment of the enterprise's financial health index more accurate, avoids the misjudgment risk that may be brought by simply relying on static indicators, improves the flexibility and adaptability of the risk warning system, and thus has higher practical value in practical applications.

[0057] The present invention solves the limitations of traditional financial prediction methods through multi-dimensional financial data collection, convolutional neural network modeling, introduction of industry correction factors, and dynamic adjustment mechanisms, improves the accuracy and efficiency of predicting the financial health status of enterprises, enhances the flexibility and adaptability of the risk warning system, and has higher practical value. BRIEF DESCRIPTION OF THE DRAWINGS

[0058] Figure 1 It is a schematic diagram of the overall method flow of the present invention.

[0059] Figure 2 This is a schematic diagram of the system module process of the present invention. Specific embodiments

[0060] To make the objectives, technical solutions and advantages of the present invention more clear and understandable, the present invention will be further described in detail below in conjunction with specific embodiments.

[0061] It should be noted that unless otherwise defined, the technical terms or scientific terms used in the present invention should have the ordinary meanings understood by those of ordinary skill in the field to which the present invention pertains. The "first", "second" and similar terms used in the present invention do not denote any order, quantity or importance, but are only used to distinguish different components. The terms such as "including" or "comprising" mean that the elements or objects appearing before this word cover the elements or objects listed after this word and their equivalents, without excluding other elements or objects. The terms such as "connected" or "coupled" are not limited to physical or mechanical connections, but may include electrical connections, whether direct or indirect. The terms such as "upper", "lower", "left" and "right" are only used to represent relative positional relationships, and when the absolute position of the object being described changes, the relative positional relationship may also change accordingly.

[0062] Embodiment:

[0063] Please refer to Figure 1 , the present invention provides a technical solution:

[0064] A method for constructing a prediction model of enterprise financial data, the specific steps include:

[0065] Step 1: Based on the enterprise financial data of the previous year's quarters of the enterprise to be predicted, extract characteristic parameter data, and the characteristic parameter data includes the enterprise's operating income, total expenditure, cost of sales, total profit, net profit, total liabilities and free cash flow;

[0066] Step 2: Preprocess the collected characteristic parameter data, and combine the operating income, total expenditure, cost of sales, total profit, net profit, total liabilities and free cash flow of the enterprise after preprocessing to generate the expense input ratio, gross profit margin, profit conversion rate and free cash flow ratio of the enterprise finance respectively. The preprocessing includes data cleaning and normalization of the collected characteristic parameter data;

[0067] Step 3: Use the generated expense input ratio, gross profit margin, profit conversion rate and free cash flow ratio of the enterprise finance to calculate the financial health index of the enterprise;

[0068] Step 4: Obtain the historical financial data of other companies in the industry to which the enterprise to be predicted belongs, calculate the revenue change rate and average profit rate of the industry to which the enterprise to be predicted belongs, and generate a financial correction factor for the industry to which the enterprise to be predicted belongs by combining the revenue change rate and the average profit rate;

[0069] Step 5: Establish a financial security prediction model, input the characteristic parameter data of the enterprise to be predicted into the financial security prediction model to obtain a financial security index, and use the calculated financial correction factor and financial health index to correct the financial security index to obtain a corrected value of the financial security index;

[0070] Step 6: Set a financial risk threshold, compare the corrected value of the financial security index with the financial risk threshold, and judge the health and security status of the enterprise's financial data.

[0071] It should be noted that data cleaning ensures the accuracy and consistency of data by removing outliers, duplicate data, and filling in missing values, avoiding the negative impact of incomplete or incorrect data on model training. Specifically, statistical methods are used to identify outliers and duplicate data, and the mean, median, or mode is used to fill in missing values, which helps to improve the quality of data and thus enhance the prediction accuracy of the model. Secondly, data normalization uses the min-max normalization method to scale all characteristic parameter data to the same scale of [0,1], eliminating the interference of different eigenvalue ranges on model training and ensuring that each characteristic has the same weight in the model training process. This not only speeds up the convergence rate of the model but also avoids certain characteristics dominating model training due to large numerical ranges, further enhancing the stability and reliability of the prediction results.

[0072] Therefore, it is necessary to perform data cleaning and normalization on the collected characteristic parameter data, and the methods are as follows:

[0073] Data cleaning of the characteristic parameter data includes the detection and deletion of outliers and duplicate data, and the handling of missing values. Specifically, statistical methods are used to identify outliers and duplicate data in the characteristic parameter data, delete the outliers and duplicate data in the characteristic parameter data, and use the mean, median, or mode of the characteristic parameter data to fill in the missing values in the characteristic parameter data;

[0074] The standardization process of the characteristic parameter data is min-max normalization, which scales the data to the range of [0,1], so that all characteristic parameter data have the same scale. The formula is as follows:

[0075]

[0076] Among them, X′ is the characteristic parameter data after normalization processing, X is the original characteristic parameter data, X minis the minimum value among the data of the same type of characteristic parameters in the dataset, X max is the maximum value among the data of the same type of characteristic parameters in the dataset.

[0077] It should be noted that the expense-income ratio, gross profit margin, profit conversion rate, and free cash flow ratio are key indicators for measuring a company's financial health and operational efficiency. The expense-income ratio reflects the company's cost control and expenditure management capabilities, and can reveal the management level in terms of expenditures; the gross profit margin measures the company's core profitability and is an important indicator for evaluating the basic profitability level of the company's products or services; the profit conversion rate reflects the efficiency of the company in converting total profit into net profit and reflects the profit quality in the operation process; the free cash flow ratio reveals the company's debt repayment ability and financial risks, and reflects the company's financial position that can be used for reinvestment or shareholder returns after meeting short-term financial obligations. Combining these indicators into a financial characteristic vector helps to comprehensively and accurately evaluate the company's financial condition, provides multi-dimensional financial health information, and thus provides strong support for financial risk prediction and decision-making.

[0078] Therefore, it is necessary to perform permutations and combinations to generate the expense input ratio, gross profit margin, profit conversion rate, and free cash flow ratio of the enterprise's finance respectively. The formulas are as follows:

[0079]

[0080]

[0081] The expense-income ratio is used to measure the company's cost management and expenditure control, the gross profit margin is used to measure the company's core profitability, the profit conversion rate is used to measure the conversion efficiency of the company from total profit to net profit, and the free cash flow ratio is used to measure the company's debt repayment ability and financial risks.

[0082] It should be noted that the company's financial health index is a comprehensive indicator for measuring the overall financial condition of the company, which can help management, investors, and analysts quickly identify the company's financial robustness and potential risks. By combining key financial indicators such as the expense-income ratio, gross profit margin, profit conversion rate, and free cash flow ratio, the financial health index can comprehensively reflect the company's profitability, cost control ability, cash flow status, and profit conversion efficiency, etc., so as to provide data support for decision-making, ensure the company's financial robustness in the fierce market competition, and promote long-term sustainable development.

[0083] Therefore, it is necessary to calculate the company's financial health index. The method is as follows:

[0084] Define the expense-income ratio, gross profit margin, profit conversion rate, and free cash flow ratio of a company's finances as the letters a, b, c, and d respectively, and comprehensively calculate the company's financial health index. The formula used is as follows:

[0085]

[0086] Among them, JK represents the company's financial health index, and a, b, c, and d represent the expense-income ratio, gross profit margin, profit conversion rate, and free cash flow ratio of the company's finances respectively. In the above formula, the larger the expense-income ratio, the smaller the financial health index, which means that the company needs to pay higher expenses for its income, which may lead to a shrinking profit margin and thus affect the overall financial health. A larger expense-income ratio usually indicates a lower financial health index, suggesting that the company may face higher operating costs and greater financial pressure. The larger the gross profit margin, profit conversion rate, and free cash flow ratio, the larger the financial health index. Among them, the larger the gross profit margin, the higher the proportion of profit retained by the company in its income, which means that the company has a stronger ability to control costs in the production or sales process and a larger profit margin, which helps to improve the financial health index. The larger the profit conversion rate, it means that the company's profit can be more effectively converted into actual net income and net profit, reflecting the excellent performance of the company's management in resource allocation and operating efficiency, and enhancing the overall financial health. The larger the free cash flow ratio, it means that after deducting the necessary capital expenditures, the company can still generate sufficient cash flow, which provides the company with more financial flexibility and risk resistance ability and can support expansion, debt repayment, and coping with sudden economic pressures.

[0087] It should be noted that the revenue change rate of the industry can reflect the revenue growth or decline trend of the entire industry in different time periods, which is very important for evaluating the company's competitive position in the industry. If the industry revenue grows rapidly, the company may be in a good market environment, otherwise it may face the risk of overall industry shrinkage. Incorporating the revenue change rate into the calculation of the correction factor helps to adjust the company's financial situation to better adapt to industry changes; the average profit margin of the industry reflects the ability of companies in the industry to convert revenue into profit. By calculating the average profit margin of multiple companies in the industry, the influence of special situations of individual companies can be eliminated, and the overall profitability of the industry can be more accurately reflected. This provides an industry standard for evaluating the company's profitability and makes the evaluation of the financial health index more objective and fair.

[0088] Therefore, combining the revenue change rate and the average profit margin, the method for generating the financial correction factor for the industry to which the company to be predicted belongs is as follows:

[0089] Obtain the annual business revenues of N other companies in the same industry in the last year and the previous year, calculate the ratio of the absolute value of the difference between the annual business revenue in the last year and the previous year to the annual business revenue in the last year, and take the average value as the revenue change rate. The formula is as follows:

[0090]

[0091] Among them, b represents the revenue change rate of the industry, are the annual business revenues of the i-th company among other companies in the last year and the previous year respectively, i represents the index of other companies in the same industry, i ∈ [1, N], and N is the number of other companies in the same industry;

[0092] Obtain the net profits and business revenues of N other companies in the same industry in the last year, calculate their ratio, and take the average value as the average profit rate. The formula is as follows:

[0093]

[0094] Among them, a represents the average profit rate of the industry, is the net profit of the i-th company among other companies in the last year;

[0095] Calculate the financial correction factor based on the collected revenue change rate and the average profit rate of the industry. The formula is as follows:

[0096]

[0097] Among them, β represents the financial correction factor; in the above formula, the larger the value of β, the better. There are two reasons for the increase in the value of β. First, the increase in the average profit rate of the industry means that companies in the industry as a whole can extract more profits from each unit of revenue. This is usually because companies in the industry perform better in aspects such as cost control, product pricing, and market positioning, or the market demand increases, resulting in a higher profit level. When the average profit rate of the industry increases, the value of β increases, indicating that the profitability of the enterprise has relatively improved, which means that the profit environment of the industry has become better, and the financial health of the enterprise in this industry may be further enhanced; Second, the decrease in the revenue change rate of the industry means that the revenue growth rate of the industry is relatively stable. Although the revenue growth does not increase significantly, the enterprise can still maintain a relatively high profit rate through effective cost control and profit improvement. Therefore, although the revenue growth of the industry slows down or is relatively stable, in the context of a lower revenue change rate, the higher profit rate makes the value of β increase, reflecting that even in an environment with slow revenue growth, the enterprise can still maintain a relatively high profit level, thereby improving its financial health index.

[0098] It should be noted that constructing an enterprise financial data prediction model and adopting a convolutional neural network structure can effectively process and learn the complex non-linear relationships in financial data, thereby improving the accuracy and robustness of financial health prediction. Through its multi-level structure, the convolutional neural network can extract deep features from the input feature vector, especially suitable for processing multi-dimensional data with spatial or temporal characteristics. In this model, the input layer receives the enterprise financial feature vector to ensure that the model can obtain comprehensive financial information; the hidden layer automatically captures the potential patterns and associations in financial data by processing data layer by layer; the output layer generates the financial health index of the enterprise to provide an accurate prediction of the enterprise's financial health status. Using the mean square error as the loss function can effectively measure the difference between the predicted value and the true value, and optimize the model parameters through the backpropagation algorithm to gradually improve the prediction accuracy of the model. This method continuously optimizes the model performance through multiple rounds of training and parameter updates to ensure high prediction reliability and stability in practical applications.

[0099] Therefore, it is necessary to establish a financial security prediction model to obtain the corrected value of the enterprise's financial security index. The method is as follows:

[0100] Establish a financial security prediction model, with the characteristic parameter data of the enterprise as the input and the financial security index of the enterprise as the output. The financial security prediction model adopts a neural network convolution structure, including an input layer, a hidden layer, and an output layer. The input layer is responsible for receiving the processed characteristic parameter data; the hidden layer is used to process the characteristic parameter data. By applying multiple convolution kernels, the hidden layer can identify complex characteristic patterns. By using the ReLu activation function, non-linear relationships are introduced to enable the model to fit complex characteristic relationships; a single independent neuron is set in the output layer, which is responsible for converting the local and high-level feature representations extracted by the hidden layer into the final prediction result;

[0101] Obtain the financial characteristic parameter data of the enterprise in historical years, and based on the evaluation and calculation by financial experts, obtain the financial security index. Use the financial characteristic parameter data of the enterprise in history as the input and the financial security index as the label to train the financial security prediction model. During the training process, select the mean square error function as the loss function, calculate the loss function value according to the output result and the true label, and calculate the gradient through the backpropagation algorithm to update the weights and biases of the neural network. Repeat the above operations until the model reaches the predetermined number of training rounds;

[0102] Then, use the calculated financial correction factor and financial health index to correct the financial security index to obtain the corrected value of the financial security index. The formula for calculating the corrected value of the financial security index is:

[0103] R safe ′=e JK+β *Rsafe

[0104] Among them, JK represents the financial health index, and R safe ′ is the correction value of the financial security index, and R safe is the financial security index; in the above formula, the reason for using the financial health index JK and the financial correction factor β to correct the financial security index R safe is that if there is a certain year when the value of R safe is very small, the values of JK and β are used to balance it. When JK increases, R safe ′ will increase appropriately, indicating that the annual revenue and profit of the enterprise company in this year are relatively good; when β increases, R safe ′ will increase appropriately, indicating that the development environment in which the enterprise is located is relatively good and has good development prospects; and the larger R safe ′ is, the better. The larger R safe ′ means the following aspects:

[0105] The profitability is enhanced. The enterprise can effectively improve core profitability indicators such as gross profit margin and net profit, indicating that its core business has strong profitability and market competitiveness;

[0106] The cost is well controlled. A higher R safe ′ value means that the enterprise performs well in cost and expense management, can effectively control operating expenses, optimize the financial structure, and avoid unnecessary expenses;

[0107] The financial risk is relatively low. With the improvement of the financial security index, the debt repayment ability of the enterprise, such as the free cash flow ratio, is enhanced, indicating that it can better cope with financial crises and external economic pressures and reduce financial risks;

[0108] The capital operation is good. A higher R safe ′ usually indicates that the enterprise's capital operation is more efficient, can maintain good cash flow, and has strong reinvestment ability to support long-term sustainable development;

[0109] The financial management is stable. Generally, the larger R safe ′ is, the more stable the enterprise's financial structure is, the more appropriate the risk management is, and it can remain stable during economic fluctuations and create better returns for shareholders and investors.

[0110] It should be noted that comparing the corrected value of the enterprise's financial security index with the financial risk threshold can effectively judge the health status of the enterprise's financial data, thereby providing timely risk warnings and decision-making support for the management. The judgment logic of Q enables the enterprise to distinguish whether there are potential financial risks based on whether the corrected value of the financial security index exceeds the set financial risk threshold. When Q = 0, it indicates that the financial data is in a healthy state and no adjustment is required; while when Q = 1, it means that the corrected value of the financial health index has exceeded the safe range, indicating that there may be risks and measures need to be taken for monitoring and adjustment. The financial risk threshold G health is set based on the analysis of the enterprise's industry characteristics, historical financial performance, and future operating environment, aiming to help the enterprise identify potential financial crises, ensure timely adoption of effective measures, avoid significant impacts of financial risks on the enterprise's operations, and safeguard the long-term healthy development of the enterprise.

[0111] Therefore, it is necessary to compare the corrected value of the financial security index output by the model with the financial risk threshold to judge the health status of the enterprise's financial data. The logical formula based on this is:

[0112]

[0113] Among them, Q represents the logical value for judging the health status of the enterprise's financial data. When Q = 0, the corrected value of the enterprise's financial security index does not exceed the financial risk threshold, indicating that there are no potential risks in the enterprise's financial data at this time and it can continue to be maintained; when Q = 1, the corrected value of the enterprise's financial security index exceeds the financial risk threshold, indicating that there are risks in the enterprise's financial data at this time and monitoring and adjustment are required; G health is the financial risk threshold.

[0114] Please refer to Figure 2 , the present invention also provides a system for constructing a model for predicting enterprise financial data. The construction system is used to execute the above-mentioned method for constructing a model for predicting enterprise financial data, including:

[0115] An enterprise financial characteristic data acquisition module, which is used to extract characteristic parameter data based on the enterprise's financial data in the previous year's quarters of the enterprise to be predicted. The characteristic parameter data includes the enterprise's operating income, total expenditure, cost of sales, total profit, net profit, total liabilities, and free cash flow;

[0116] Feature data preprocessing module, which is used to preprocess the collected feature parameter data, and combine the operating income, total expenditure, cost of sales, total profit, net profit, total liabilities and free cash flow of the enterprise after preprocessing to generate the expense input ratio, gross profit margin, profit conversion rate and free cash flow ratio of the enterprise's finance respectively. The preprocessing includes data cleaning and normalization of the collected feature parameter data;

[0117] Financial health index generation module, which is used to calculate the financial health index of the enterprise by using the generated expense input ratio, gross profit margin, profit conversion rate and free cash flow ratio of the enterprise's finance;

[0118] Financial correction factor generation module, which is used to obtain the historical financial data of other enterprise companies in the industry to which the enterprise to be predicted belongs, calculate the revenue change rate and average profit rate of the industry to which the enterprise to be predicted belongs, and generate the financial correction factor of the industry to which the enterprise to be predicted belongs in combination with the revenue change rate and average profit rate;

[0119] Financial security prediction model construction module, which is used to establish a financial security prediction model, input the feature parameter data of the enterprise to be predicted into the financial security prediction model to obtain the financial security index, and use the calculated financial correction factor and financial health index to correct the financial security index to obtain the corrected value of the financial security index;

[0120] Risk warning module, which is used to set a financial risk threshold, compare the corrected value of the financial security index with the financial risk threshold, and judge the health and safety status of the enterprise's financial data.

[0121] The above formulas are all dimensionless and take their numerical values for calculation. The formulas are obtained by software simulation of a large amount of collected data to get a formula closest to the real situation. The preset parameters in the formulas are set by those skilled in the art according to the actual situation.

[0122] The above embodiments can be implemented in whole or in part by software, hardware, firmware or any other combination. When implemented using software, the above embodiments can be implemented in whole or in part in the form of a computer program product. Those skilled in the art can realize that the units and algorithm steps of each example described in combination with the embodiments disclosed herein can be implemented by electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are executed by hardware or software methods depends on the specific application and design constraints of the technical solution.

[0123] The unit described as a separation component may or may not be physically separated. The component shown as a unit may or may not be a physical unit, and it may be located in one place or distributed across multiple network units. Some or all of the units can be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0124] As described above, the foregoing is only a specific implementation manner of this application, but the protection scope of this application is not limited thereto. Any person skilled in the art within the technical scope disclosed in this application can easily think of changes or substitutions, which should all be covered within the protection scope of this application.

Claims

1. A method for constructing a model for predicting enterprise financial data, characterized in that: The specific steps include: Step 1: Extract characteristic parameter data based on the enterprise financial data of the previous quarter of the enterprise to be predicted, wherein the characteristic parameter data includes the enterprise's operating income, total expenditure, sales cost, total profit, net profit, total debt and free cash flow; Step 2: Preprocessing the collected characteristic parameter data, combining the preprocessed business income, total expenditure, sales cost, total profit, net profit, total liabilities and free cash flow of the enterprise, and generating the cost input ratio, gross profit margin, profit conversion rate and free cash flow ratio of the enterprise finance respectively, wherein the preprocessing includes data cleaning and normalization of the collected characteristic parameter data; Step 3: Calculate the financial health index of the enterprise using the generated cost input ratio, gross profit margin, profit conversion ratio and free cash flow ratio of the enterprise financials; Step 4: Obtain the historical financial data of other companies in the industry to which the enterprise to be predicted belongs, calculate the income change rate and average profit rate of the industry to which the enterprise to be predicted belongs, and generate the financial correction factor of the industry to which the enterprise to be predicted belongs by combining the income change rate and the average profit rate; Step 5: Establish a financial security prediction model, input the characteristic parameter data of the enterprise to be predicted into the financial security prediction model to obtain a financial security index, and use the calculated financial correction factor and financial health index to correct the financial security index to obtain a financial security index correction value; Step 6: Establish a financial risk threshold, compare the revised value of the financial security index with the financial risk threshold, and determine the health and safety status of the company's financial data; The collected characteristic parameter data is cleaned and normalized according to the following method: Data cleaning of characteristic parameter data includes detection and deletion of outliers and duplicate data, and processing of missing values. Specifically, it uses statistical methods to identify outliers and duplicate data in characteristic parameter data, deletes outliers and duplicate data in characteristic parameter data, and uses the mean, median or mode of characteristic parameter data to fill in missing values ​​in characteristic parameter data; The characteristic parameter data is normalized as minimum-maximum normalization, scaling the data to the range of [0,1] so that all characteristic parameter data have the same scale. The formula is: Among them, X′ is the normalized feature parameter data, X is the original feature parameter data, and X min is the minimum value of the same type of feature parameter data in the data set, X max It is the maximum value of the same type of feature parameter data in the data set; The formulas for generating the cost input ratio, gross profit margin, profit conversion rate and free cash flow ratio of the enterprise finance are as follows: The expense-to-income ratio is used to measure the cost management and expense control of an enterprise, the gross profit margin is used to measure the core profitability of an enterprise, the profit conversion rate is used to measure the efficiency of the enterprise's conversion from gross profit to net profit, and the free cash flow ratio is used to measure the debt-paying ability and financial risk of an enterprise; The method for calculating the financial health index of an enterprise is based on: Define the cost-to-income ratio, gross profit margin, profit conversion rate and free cash flow ratio of the enterprise's finance as letters a, b, c and d respectively, and comprehensively calculate the enterprise's financial health index based on the formula: Among them, JK represents the financial health index of the enterprise, a, b, c, d represent the cost-to-income ratio, gross profit margin, profit conversion rate and free cash flow ratio of the enterprise's finance respectively; Combining the income change rate and the average profit margin, the financial correction factor of the industry to which the enterprise to be predicted belongs is generated based on the following method: Obtain the annual operating income of last year and the previous year of N other companies in the same industry, calculate the ratio of the absolute value of the difference between the annual operating income of last year and the previous year to the annual operating income of last year, and take the average value as the income change rate. The formula is: Among them, b represents the income change rate of the industry, are the annual operating income of the i-th company of other enterprises last year and the annual operating income of the previous year, respectively. i represents the index of other companies in the same industry, i∈[1,N], and N is the number of other companies in the same industry. Obtain the net profit and operating income of N other companies in the same industry last year, calculate their ratio, and take the average as the average profit rate. The formula is: Among them, a represents the average profit margin of the industry, is the net profit of the i-th company of other enterprises last year; The financial correction factor is calculated based on the rate of change of collected revenue and the average profit margin of the industry, based on the formula: Among them, β represents the financial correction factor; A financial security prediction model is established to obtain the revised value of the enterprise's financial security index, based on the following method: A financial security prediction model is established, with the characteristic parameter data of the enterprise as input and the financial security index of the enterprise as output. The financial security prediction model adopts a neural network convolution structure, including an input layer, a hidden layer and an output layer. The input layer is responsible for receiving the processed characteristic parameter data; the hidden layer is used to process the characteristic parameter data. By applying multiple convolution kernels, the hidden layer can recognize complex characteristic patterns. By using the ReLu activation function, nonlinear relationships are introduced so that the model can fit complex characteristic relationships; an independent neuron is set in the output layer, which is responsible for converting the local and high-level feature representations extracted by the hidden layer into the final prediction results; The financial correction factor and financial health index calculated are then used to correct the financial security index to obtain the corrected value of the financial security index. The formula for calculating the corrected value of the financial security index is: R safe ′=e JK+β *R safe Among them, JK represents the financial health index, R safe ′ is the revised value of financial security index, R safe is the financial security index; Compare the revised value of the financial security index output by the model with the financial risk threshold to determine the health status of the company's financial data. The logical formula is: Among them, Q represents the logical value for judging the health status of the enterprise's financial data. When Q = 0, the revised value of the enterprise's financial security index does not exceed the financial risk threshold, indicating that there is no potential risk in the enterprise's financial data at this time, and it can be maintained; when Q = 1, the revised value of the enterprise's financial security index exceeds the financial risk threshold, indicating that there is a risk in the enterprise's financial data at this time, and monitoring and adjustment are required; G health is the financial risk threshold.

2. A system for building a model for predicting corporate financial data, characterized in that: The construction system is used to execute the method for constructing a forecast enterprise financial data model according to any one of claim 1, comprising: An enterprise financial characteristic data collection module, which is used to extract characteristic parameter data based on the enterprise financial data of the enterprise to be predicted in the previous quarter of the previous year, and the characteristic parameter data includes the enterprise's operating income, total expenditure, sales cost, total profit, net profit, total liabilities and free cash flow; A feature data preprocessing module, which is used to preprocess the collected feature parameter data, and to generate the cost input ratio, gross profit margin, profit conversion rate and free cash flow ratio of the enterprise's finance by permutation and combination according to the preprocessed business income, total overhead expenditure, sales cost, total profit, net profit, total liabilities and free cash flow. The preprocessing includes data cleaning and normalization of the collected feature parameter data; A financial health index generation module, the financial health index generation module is used to calculate the financial health index of the enterprise by using the generated cost input ratio, gross profit margin, profit conversion rate and free cash flow ratio of the enterprise finance; A financial correction factor generation module, which is used to obtain historical financial data of other companies in the industry to which the enterprise to be predicted belongs, so as to calculate the income change rate and average profit rate of the industry to which the enterprise to be predicted belongs, and to generate a financial correction factor for the industry to which the enterprise to be predicted belongs by combining the income change rate and the average profit rate; Constructing a financial security prediction model module, the financial security prediction model module is used to establish a financial security prediction model, inputting characteristic parameter data of the enterprise to be predicted into the financial security prediction model to obtain a financial security index, and using the calculated financial correction factor and financial health index to correct the financial security index to obtain a financial security index correction value; The risk warning module is used to establish a financial risk threshold, compare the financial security index correction value with the financial risk threshold, and determine the health and safety status of the company's financial data.

Images