Enterprise account fund centralized collection degree calculation method and device

Abstract

The invention discloses an enterprise account fund centralized collection degree calculation method and device, relates to the technical field of big data, and can also be used in the financial field, and the method comprises the steps: pre-constructing a fund concentration degree collection model for fund flow data of a plurality of business systems in an enterprise; performing collection analysis on the input fund flow data based on the fund concentration degree collection model to obtain a calculation result of a fund concentration degree; and determining an evaluation result of the current fund collection state of the enterprise according to the calculation result of the fund centralized collection degree. According to the invention, comprehensive capture and accurate analysis of the enterprise fund collection influence factors are realized. By optimizing the data processing and model training process, the accuracy and comprehensiveness of the fund centralized collection degree model are improved, and powerful support is provided for enterprise fund management and decision making.

Description

Technical Field

[0001] The present invention relates to the field of big data technology and can also be used in the financial field, and in particular to a method and device for calculating the concentration degree of funds in enterprise accounts. Background Art

[0002] In the context of corporate fund management, in order to further improve the operational efficiency of funds and prevent and control fund risks, some companies use the enterprise-level treasury platform to aggregate the business data of the group and its member units to generate a fund concentration degree. The fund concentration degree reflects the degree of collection and concentration of corporate funds and is usually used as a key indicator to measure the efficiency and utilization rate of corporate fund management. Generally speaking, the higher the fund concentration, the higher the efficiency of corporate fund management, which can better utilize existing funds and improve the overall fund operation effect.

[0003] However, the current fund concentration model has many limitations. In the process of fund concentration, the systems involved within the enterprise are often very complex, and the systems used by different subsidiaries are also significantly different. In addition, the quality of business data reporting varies, and the rules for data preprocessing are inconsistent, resulting in the inability to conduct in-depth multi-dimensional analysis of the flow of account amounts. Traditional models usually rely on single-dimensional statistical indicators, with relatively simple rules, which are prone to data distortion, resulting in low accuracy of the concentration model.

[0004] In other words, it is difficult to comprehensively evaluate the fund management status of corporate accounts with a single statistical method. Such limitations make it impossible for companies to scientifically and accurately identify potential risks when conducting risk assessments and forecasting fund flows, which may lead to problems such as idle or insufficient funds. Poor fund management will not only affect the liquidity of the company, but may also seriously reduce its overall operating efficiency and competitiveness.

[0005] This section is intended to provide a background or context to the embodiments of the invention recited in the claims. No admission is made that the description herein is prior art by inclusion in this section. Summary of the invention

[0006] The embodiment of the present invention provides a method and device for calculating the concentration degree of funds in an enterprise account, which are used to solve the problems in the prior art.

[0007] An embodiment of the present invention provides a method for calculating the concentration degree of funds in an enterprise account, including:

[0008] A fund concentration aggregation model pre-constructed from fund flow data of multiple business systems in an enterprise, wherein the aggregation model is obtained by standardizing and reducing the dimension of the collected fund flow data through principal component analysis, and the fund flow data includes account balances, accounts receivable and payable information, financing information, bill information and settlement information;

[0009] Based on the fund concentration degree collection model, the input fund flow data is collected and analyzed to obtain a calculation result of the fund concentration degree, wherein the calculation result represents the fund factors that affect the current fund concentration degree;

[0010] An evaluation result of the current fund collection status of the enterprise is determined according to the calculation result of the fund concentration degree, and the evaluation result includes the fund collection status of multiple business systems, risk warning and fund optimization suggestions.

[0011] According to any embodiment of the present invention, the process of constructing the fund concentration model includes:

[0012] Collect fund flow data from multiple business systems through an enterprise-level treasury platform;

[0013] Preprocessing the collected capital flow data and storing it in the data center, wherein the preprocessing includes data cleaning, data conversion, data standardization, data integration and data compression;

[0014] The principal component analysis is used to analyze the main factors affecting the fund concentration in the fund flow data of the data center to construct the fund concentration model.

[0015] According to any embodiment of the present invention, determining the main factors affecting the fund concentration in the fund flow data of the data center through principal component analysis and constructing the fund concentration model includes:

[0016] Identify the key factors that affect the degree of fund concentration through the fund flow data of the data center, and calculate the fund index data according to the key factors, including the uncollected funds within the attributable scope, the collected funds within the attributable scope, the uncollected funds within the full scope, the collected funds within the full scope, the structure analysis of the uncollected funds, the total amount of funds, the total amount of funds that can be collected, and the total amount of funds that cannot be collected within the attributable scope;

[0017] The collected funding indicator data were standardized using principal component analysis to ensure that the weights of each feature in the analysis were consistent;

[0018] The multi-dimensional capital index data is converted into low-dimensional principal component data using principal component analysis, and the capital concentration aggregation model is constructed based on the principal component data.

[0019] According to any embodiment of the present invention, the use of principal component analysis to standardize the collected capital indicator data to ensure that the weights of various features in the analysis are consistent includes:

[0020] Calculate the mean and standard deviation of the collected capital indicator data to obtain the average level and dispersion of the capital indicator data respectively;

[0021] Each data point is converted into standardized data based on a standardization formula so that the standardized data will have distribution characteristics with a mean of zero and a standard deviation of one.

[0022] According to any embodiment of the present invention, the method of converting multi-dimensional fund indicator data into low-dimensional principal component data using principal component analysis includes:

[0023] Calculate the covariance between data sets consisting of multiple fund indicator data to determine the volatility of the fund concentration in each indicator;

[0024] Use principal component analysis to determine the direction with the largest variance in the fund indicator data, and retain the principal components with variance ratios exceeding the preset value;

[0025] According to the amount of information contained in the principal components, the principal components with higher cumulative variance ratios are retained and the secondary components with smaller cumulative variance ratios are removed to achieve principal component dimensionality reduction.

[0026] According to any embodiment of the present invention, constructing the fund concentration model according to the principal component data includes:

[0027] Using the principal component data after dimensionality reduction as input features and the fund concentration degree as the target variable, a multivariate linear regression model of fund concentration degree is constructed;

[0028] The model parameters of the multivariate linear regression model are solved by the least square method, wherein the model parameters are grouped according to the dimension conditions of time and enterprise code, and are aggregated on the dimension of the only enterprise group.

[0029] According to any embodiment of the present invention, the step of determining the evaluation result of the current fund collection status of the enterprise according to the calculation result of the fund concentration degree includes:

[0030] Determining the fund collection efficiency of multiple business systems according to the calculation result of the fund concentration degree;

[0031] Based on the results of fund collection efficiency analysis, for business systems with low collection rates, risk warnings are generated based on the fund flow and fluctuations of each system;

[0032] Generate optimization suggestions for enterprise fund collection based on the fund collection efficiency analysis results and generated risk warnings.

[0033] The embodiment of the present invention also provides a device for calculating the concentration degree of funds in enterprise accounts, which is used to solve the problems in the prior art. The device includes:

[0034] A data input module is used to pre-construct a fund concentration aggregation model for fund flow data of multiple business systems in an enterprise, wherein the aggregation model is obtained by standardizing and reducing the dimension of the collected fund flow data through principal component analysis, and the fund flow data includes account balances, receivables and payables information, financing information, bill information and settlement information;

[0035] A concentration degree calculation module, used for performing concentration analysis on the input capital flow data based on the capital concentration degree concentration model to obtain a calculation result of the capital concentration degree, wherein the calculation result represents the capital factors affecting the current capital concentration degree;

[0036] The concentration degree evaluation module is used to determine the evaluation result of the enterprise's current fund collection status according to the calculation result of the fund concentration degree, and the evaluation result includes the fund collection status of multiple business systems, risk warning and fund optimization suggestions.

[0037] According to any embodiment of the present invention, the process of constructing the fund concentration model includes:

[0038] Data collection module, used to collect fund flow data from multiple business systems through the enterprise-level treasury platform;

[0039] A data preprocessing module, used to preprocess the collected capital flow data and store it in the data center, wherein the preprocessing includes data cleaning, data conversion, data standardization, data integration and data compression;

[0040] The model building module is used to analyze the main factors affecting the concentration of funds in the fund flow data of the data center through principal component analysis to build the fund concentration model.

[0041] According to any embodiment of the present invention, the model building module includes:

[0042] An indicator determination unit is used to identify key factors affecting the degree of fund concentration through the fund flow data of the data center, and calculate fund indicator data according to the key factors, including uncollected funds within the attributable scope, collected funds within the attributable scope, uncollected funds within the full scope, collected funds within the full scope, uncollected fund structure analysis, total funds, total collectable funds, and total uncollected funds within the attributable scope;

[0043] A standardization processing unit is used to standardize the collected capital indicator data using principal component analysis to ensure that the weights of each feature in the analysis are consistent;

[0044] The data dimension reduction unit is used to convert the multi-dimensional capital indicator data into low-dimensional principal component data using principal component analysis, and to construct the capital concentration aggregation model based on the principal component data.

[0045] According to any embodiment of the present invention, the standardization processing unit is specifically used for:

[0046] Calculate the mean and standard deviation of the collected capital indicator data to obtain the average level and dispersion of the capital indicator data respectively;

[0047] Each data point is converted into standardized data based on a standardization formula so that the standardized data will have distribution characteristics with a mean of zero and a standard deviation of one.

[0048] According to any embodiment of the present invention, the data dimension reduction unit is specifically used for:

[0049] Calculate the covariance between data sets consisting of multiple fund indicator data to determine the volatility of the fund concentration in each indicator;

[0050] Use principal component analysis to determine the direction with the largest variance in the fund indicator data, and retain the principal components with variance ratios exceeding the preset value;

[0051] According to the amount of information contained in the principal components, the principal components with higher cumulative variance ratios are retained and the secondary components with smaller cumulative variance ratios are removed to achieve principal component dimensionality reduction.

[0052] According to any embodiment of the present invention, the data dimension reduction unit is further used for:

[0053] Using the principal component data after dimensionality reduction as input features and the fund concentration degree as the target variable, a multivariate linear regression model of fund concentration degree is constructed;

[0054] The model parameters of the multivariate linear regression model are solved by the least square method, wherein the model parameters are grouped according to the dimension conditions of time and enterprise code, and are aggregated on the dimension of the only enterprise group.

[0055] According to any embodiment of the present invention, the aggregation degree evaluation module includes:

[0056] A collection efficiency determination unit, used to determine the fund collection efficiency of multiple business systems according to the calculation result of the fund concentration degree;

[0057] The risk warning unit is used to generate risk warnings based on the results of fund collection efficiency analysis for business systems with low collection rates, combined with the fund flows and fluctuations of each system;

[0058] The optimization suggestion unit is used to generate optimization suggestions for the enterprise's fund collection based on the fund collection efficiency analysis results and the generated risk warnings.

[0059] An embodiment of the present invention also provides a computer device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor implements the above-mentioned enterprise account fund concentration degree calculation method when executing the computer program.

[0060] An embodiment of the present invention further provides a computer-readable storage medium, wherein the computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the above-mentioned method for calculating the concentration degree of funds in an enterprise account is implemented.

[0061] An embodiment of the present invention further provides a computer program product, which includes a computer program. When the computer program is executed by a processor, the above-mentioned method for calculating the concentration degree of funds in an enterprise account is implemented.

[0062] The enterprise account fund concentration calculation method and device provided by the embodiment of the present invention realizes the comprehensive capture and accurate analysis of the factors affecting the enterprise fund concentration. By optimizing the data processing and model training process, the accuracy and comprehensiveness of the fund concentration model are improved, providing strong support for enterprise fund management and decision-making. BRIEF DESCRIPTION OF THE DRAWINGS

[0063] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments or the prior art descriptions. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work. In the drawings:

[0064] Figure 1 This is one of the flow charts of the method for calculating the concentration degree of enterprise account funds in an embodiment of the present invention;

[0065] Figure 2 This is a second flow chart of a method for calculating the concentration degree of enterprise account funds in an embodiment of the present invention;

[0066] Figure 3 This is a third flow chart of a method for calculating the concentration degree of enterprise account funds in an embodiment of the present invention;

[0067] Figure 4 This is a fourth flow chart of the method for calculating the concentration degree of enterprise account funds in an embodiment of the present invention;

[0068] Figure 5This is a fifth flow chart of the method for calculating the concentration degree of enterprise account funds in an embodiment of the present invention;

[0069] Figure 6 This is a sixth flow chart of the method for calculating the concentration degree of enterprise account funds in an embodiment of the present invention;

[0070] Figure 7 This is a seventh flow chart of a method for calculating the concentration degree of enterprise account funds in an embodiment of the present invention;

[0071] Figure 8 This is a flowchart of the method for calculating the concentration degree of enterprise account funds in an embodiment of the present invention;

[0072] Fig. 9 This is one of the structural diagrams of the enterprise account funds concentration degree calculation device in an embodiment of the present invention;

[0073] Fig.10 Schematic diagram of the structure of an electronic device in an embodiment of the present invention. DETAILED DESCRIPTION

[0074] To make the purpose, technical solution and advantages of the embodiments of the present invention more clear, the embodiments of the present invention are further described in detail below in conjunction with the accompanying drawings. Here, the exemplary embodiments of the present invention and their descriptions are used to explain the present invention, but are not intended to limit the present invention.

[0075] The information collected in the technical solution of the present invention is information and data authorized by the user or fully authorized by all parties, and the collection, storage, use, processing, transmission, provision, disclosure and application of the relevant data comply with the relevant laws, regulations and standards of the relevant countries and regions, take necessary confidentiality measures, do not violate public order and good morals, and provide corresponding operation entrances for users to choose to authorize or refuse.

[0076] The acquisition, transmission, storage, use, and processing of data in the technical solution of the present invention are in compliance with the relevant provisions of national laws and regulations.

[0077] It should be noted that in the embodiments of the present invention, certain software, components, models and other existing solutions in the industry may be mentioned, which should be regarded as exemplary. Their purpose is only to illustrate the feasibility of implementing the technical solution of the present invention, but it does not mean that the applicant has or will necessarily use the solution.

[0078] However, considering the limitations of the traditional fund concentration degree model, the present invention provides a method and device for calculating the concentration degree of enterprise account funds, which can fully capture and accurately analyze the factors affecting enterprise fund concentration. By optimizing the data processing and model training process, the accuracy and comprehensiveness of the fund concentration degree model are improved, providing strong support for enterprise fund management and decision-making.

[0079] The present invention provides an embodiment of a method for calculating the concentration degree of funds in an enterprise account, see Figure 1 The method for calculating the concentration degree of enterprise account funds specifically includes the following contents:

[0080] Step S11: A funds concentration aggregation model is pre-constructed based on the funds flow data of multiple business systems in the enterprise, wherein the aggregation model is obtained by standardizing and reducing the dimension of the collected funds flow data through principal component analysis, and the funds flow data includes account balances, accounts receivable and payable information, financing information, bill information and settlement information.

[0081] First, the capital flow data from multiple business systems of the enterprise is input into the pre-built capital concentration aggregation model, which is designed to analyze and evaluate the overall capital flow status of the enterprise. Data types include account balances, accounts receivable and payable information, financing information, bill circulation and settlement information, etc., which can fully reflect the use and flow of funds within the enterprise. By integrating and inputting these data into the model, a global perspective of capital flow can be established, which is convenient for subsequent concentration analysis.

[0082] Based on the input data, the model uses principal component analysis (PCA) to standardize and reduce the dimension of the capital flow data. Standardization ensures that data from different systems have a unified measurement standard in the analysis, making the data comparable. Dimensionality reduction is to filter out the principal component that has the greatest impact on capital concentration, eliminate redundant information, reduce model complexity, and retain the main features of the data.

[0083] Step S12: Based on the fund concentration degree aggregation model, the input fund flow data is aggregated and analyzed to obtain a calculation result of the fund concentration degree, wherein the calculation result represents the fund factors that affect the current fund concentration degree.

[0084] The model collects and analyzes the standardized and reduced-dimensional capital flow data to generate the calculation results of capital concentration. This result specifically quantifies the capital concentration of each account and business system of the enterprise, reflecting the overall performance of the current centralized capital management of the enterprise. By analyzing the calculation results of the concentration degree, the key factors affecting the capital concentration of the enterprise can be revealed, such as the uneven distribution of funds or liquidity differences in each business system, providing insights into potential problems in the capital management of the enterprise.

[0085] Step S13: Determine the evaluation result of the current fund collection status of the enterprise according to the calculation result of the fund concentration degree, and the evaluation result includes the fund collection status of multiple business systems, risk warning and fund optimization suggestions.

[0086] Based on the calculation results of fund concentration, enterprises can clearly understand the current fund collection status. The evaluation not only provides a detailed analysis of the fund collection status of each business system, but also points out the potential risks in fund management. For example, business systems with low concentration may have the risk of fund dispersion or inefficient utilization. In addition, the report will also generate corresponding risk warning prompts to help enterprise management identify potential fund risks in advance. At the same time, fund optimization suggestions will be put forward based on the analysis results, such as adjusting fund collection strategies or improving fund flows in specific business systems, in order to improve the overall fund management efficiency of the enterprise.

[0087] From the above description, it can be seen that the enterprise account fund concentration calculation method provided by the embodiment of the present invention can realize the comprehensive capture and accurate analysis of the factors affecting the enterprise fund concentration. By optimizing the data processing and model training process, the accuracy and comprehensiveness of the fund concentration model are improved, providing strong support for enterprise fund management and decision-making.

[0088] In one embodiment of the method for calculating the concentration degree of enterprise account funds of the present invention, see Figure 2 The construction process of the fund concentration aggregation model includes:

[0089] Step S1: Collect fund flow data from multiple business systems through the enterprise-level treasury platform.

[0090] A common problem faced by enterprises is that each business system is relatively independent, resulting in the existence of "information islands". Different departments use inconsistent equipment systems, various business systems are scattered, and data cannot be centrally managed, resulting in low efficiency in the use of enterprise data and information resources, affecting overall capital management and operational efficiency.

[0091] Therefore, enterprises build enterprise-level treasury platforms to uniformly manage the data of various business systems. In terms of specific implementation, data collection components such as Kettle are used to extract, transform and load (ETL) data from multiple business systems. These business systems cover various important modules of the enterprise, including account management, settlement, budget, basic finance, bills, financing, fund management, bank-enterprise direct connection, guarantee management, and accounts receivable and payable management. By extracting data from these different systems and loading them uniformly into the treasury center, centralized storage and management of multi-source data can be achieved.

[0092] Furthermore, in addition to Kettle, the present invention does not limit other data collection tools, such as Apache Sqoop or DataX. These tools also have powerful data collection and conversion capabilities, and can be flexibly configured and used according to the specific needs of the enterprise. By reasonably selecting and configuring these tools, data integration can be performed more efficiently, and unified management of capital flow data can be achieved, laying a good data foundation for subsequent data analysis and capital concentration model construction.

[0093] Step S2: Preprocess the collected capital flow data and store it in the data center, wherein the preprocessing includes data cleaning, data conversion, data standardization, data integration and data compression.

[0094] Before building a fund concentration model, enterprises need to preprocess the collected initial data. The main purpose of preprocessing is to improve the quality of data to ensure the accuracy of model analysis, reduce data deviation, and enhance the reliability of the model. The preprocessing process is divided into the following key steps:

[0095] 1. Data cleaning:

[0096] First, the collected data needs to be processed initially to remove records that are irrelevant or invalid to the modeling. For missing values, statistical methods such as mean, median or mode can be used to fill them. For some important fields, the missing data can also be interpolated through the prediction model to ensure the integrity of the data.

[0097] 2. Data conversion:

[0098] Since the data formats of different systems may be inconsistent, all data need to be converted into a unified format. This conversion can reduce the differences between features, make subsequent processing smoother, and ensure that data from different sources are comparable.

[0099] 3. Data Normalization:

[0100] Through data normalization, we ensure that all data follows a unified conceptual model and maintains data consistency and accuracy. Data normalization helps eliminate outliers and improves the stability of subsequent analysis.

[0101] 4. Data Integration and Analysis:

[0102] In the data integration phase, data from different data sources need to be grouped, aggregated, and analyzed according to business logic. Enterprises analyze this data by module and build a topic model to ensure the validity of the data. Feature selection is also an important step in this phase. By selecting the features that have the greatest impact on model prediction, the computing cost can be effectively reduced and the prediction accuracy of the model can be improved.

[0103] 5. Data Storage:

[0104] Finally, the integrated and processed data needs to be stored efficiently. Enterprises usually use Orc+Snappy column compression technology for storage to reduce storage space usage and increase data query and analysis speed. At the same time, the unprocessed original data will also be backed up to the source area for possible subsequent tracing.

[0105] For example, in the selection of data architecture, enterprises can choose appropriate technical solutions according to their own needs. Hadoop architecture is suitable for processing large-scale data sets and has the advantages of distributed storage and computing. Mpp architecture focuses more on parallel processing and query performance, and is suitable for scenarios with high requirements for real-time and high performance. Enterprises can choose appropriate data storage and processing architecture according to their business characteristics and technical requirements to ensure the efficiency and flexibility of data processing.

[0106] Step S3: Analyze the main factors affecting the fund concentration in the fund flow data of the data center through principal component analysis to construct the fund concentration model.

[0107] Among them, after the pre-processed fund flow data is stored in the data center, the present invention uses principal component analysis (PCA) to further analyze these data. Principal component analysis aims to extract the key factors that have the greatest impact on fund concentration from a large amount of fund flow data. By identifying the most representative principal components, the model can eliminate redundant information and reduce the dimensions of analysis. It not only simplifies the complexity of the data, but also retains the most important variables in the calculation of fund concentration, such as changes in account balances, fund liquidity, and accounts receivable and payable turnover. Finally, the fund concentration model is constructed through the analysis results, which can accurately reflect the fund concentration status of each system of the enterprise and provide effective support for the enterprise's fund management decisions.

[0108] In one embodiment of the method for calculating the concentration degree of enterprise account funds of the present invention, see Figure 3 The method of determining the main factors affecting the fund concentration degree in the fund flow data of the data center through principal component analysis and constructing the fund concentration degree model includes:

[0109] Step S3A: Identify the key factors affecting the degree of fund concentration through the fund flow data of the data center, and calculate the fund indicator data according to the key factors, including uncollected funds within the attributable scope, collected funds within the attributable scope, uncollected funds within the full scope, collected funds within the full scope, uncollected funds structure analysis, total funds, total collectable funds and total uncollected funds within the attributable scope.

[0110] Among them, based on the data integration of the data center and through business combing, the present invention identifies multiple channels that affect account funds, including bills, settlements, accounts payable and receivable, financing, bank deposits, loan contracts, and financing guarantees. Based on the above factors, the present invention further derives the parameter calculation formula of the downstream fund collection index as follows:

[0111] Uncollected funds that can be classified into the following categories: bank account balance minus (the balance of an account with a security deposit, frozen funds, special account or fiscal appropriation on the statistical date).

[0112] Funds collected that can be attributed to: the balance of each member unit in the finance company’s account.

[0113] Full-scope uncollected funds: the balance of each member unit’s bank account.

[0114] Full-scope collected funds: the balance of each member unit in the finance company’s account.

[0115] The present invention further refines the analysis of the uncollected funds structure into the following two aspects:

[0116] Total amount of funds (total amount of restricted funds): the total balance of accounts with the account types of margin, frozen funds, special accounts or fiscal appropriations on the statistical date.

[0117] Total amount of funds that can be collected: bank account balance minus (the balance of accounts with account types of deposit, frozen funds, special accounts or fiscal appropriations on the statistical date).

[0118] The total amount of uncollected funds that can be classified is: the bank account balance minus (the balance of accounts with account types of deposit, frozen funds, special accounts or fiscal appropriations on the statistical date).

[0119] Step S3B: Use principal component analysis to standardize the collected funding indicator data to ensure that the weights of each feature in the analysis are consistent.

[0120] After obtaining the key indicators of capital flow, the present invention standardizes these data through principal component analysis. The purpose of the standardization is to ensure that different features have consistent weights so that in subsequent analysis, each indicator can be compared under the same standard. In this way, the scale differences between the data are eliminated, allowing the model to fairly evaluate the influence of each capital indicator. The standardized data ensures that the model can more accurately reflect the true role of each capital factor when calculating the capital concentration degree.

[0121] In an alternative embodiment, see Figure 4 , the principal component analysis is used to standardize the collected capital indicator data to ensure that the weights of each feature in the analysis are consistent, including:

[0122] Step S3B1: Calculate the mean and standard deviation of the collected capital index data to obtain the average level and dispersion of the capital index data respectively.

[0123] Step S3B2: Convert each data point into standardized data based on a standardized formula, so that the standardized data will have a distribution characteristic with a mean of zero and a standard deviation of one.

[0124] The present invention uses PCA to convert the indicators that affect the fund concentration into new data with a mean of 0 and a standard deviation of 1. This step ensures that each feature (or variable) has the same weight in the PCA analysis, as shown in Formula 1:

[0125]

[0126] Among them, x refers to the balance indicator data used to measure the centralized collection of account funds, which is specifically expressed in the form of a two-dimensional array. μ represents the mean of the feature data, that is, the average level of all values ​​in x. And σ represents the standard deviation of the feature, which is used to measure the degree of dispersion of data distribution.

[0127] After the normalization process, the data z will present a distribution with a mean of 0 and a standard deviation of 1. This conversion process is based on a specific mathematical formula to calculate the Z score of each data point. The Z score reflects the relative position of the data point relative to the mean and standard deviation of the distribution in which it is located, which helps the present invention to analyze and interpret the data more accurately.

[0128] Step S3C: Use principal component analysis to convert the multi-dimensional capital indicator data into low-dimensional principal component data, and construct the capital concentration aggregation model based on the principal component data.

[0129] After standardization, the capital data may still contain many dimensions, which will increase the complexity of the model. Therefore, another key step of the principal component analysis technology is to convert the multi-dimensional capital indicator data into low-dimensional principal component data. By screening out the principal components that have the greatest impact on the concentration of capital, the model can remove redundant data, simplify calculations, and focus on analyzing the most important capital factors. This not only improves the analysis efficiency, but also ensures the accuracy and predictive ability of the model.

[0130] In an alternative embodiment, see Figure 5 , the use of principal component analysis to convert multi-dimensional capital indicator data into low-dimensional principal component data includes:

[0131] Step S3C1: Calculate the covariance between data sets consisting of multiple fund indicator data to determine the volatility of the fund concentration in each indicator.

[0132] Step S3C2: Use principal component analysis to determine the direction with the largest variance in the fund indicator data, and retain the principal component with a variance ratio exceeding a preset value.

[0133] Step S3C3: According to the amount of information contained in the principal component, retain the principal component with a higher cumulative variance ratio and remove the secondary component with a smaller cumulative variance ratio to achieve principal component dimensionality reduction.

[0134] Since PCA is an unsupervised learning method, the target variable (such as the total amount of unclassified funds) will not be considered in the dimensionality reduction process. This step is completely dependent on the aforementioned data standardization results.

[0135] The present invention needs to convert the index data related to the fund concentration degree into a set of new variables, namely principal components (PCs), through linear transformation. These new variables are linear combinations of the original variables. The present invention achieves this by finding the direction with the largest variance in the data through PCA.

[0136] For a data set containing (n) random variables, the (i, j)th element of its covariance matrix is ​​the covariance between the random variables (X_i) and (X_j), see formula (2):

[0137]

[0138] where (E[X_i]) and (E[X_j]) are the means of (X_i) and (X_j) respectively.

[0139] The diagonal elements of the covariance matrix, i.e. the variance of each fund concentration amount indicator, accurately reflect the fund collection volatility of the indicator. Similarly, the diagonal elements of the covariance matrix also represent the variance of each fund collection indicator, effectively measuring the volatility of the indicator.

[0140] The off-diagonal elements of the covariance matrix represent the covariance between the two financial indicators, which reveals the strength and direction of the linear relationship between them. A positive value indicates a positive correlation between the two, while a negative value indicates a negative correlation between them.

[0141] In addition, the relative size of the amount of information contained in each principal component can be quantified by its variance ratio. The larger the variance ratio, the richer the amount of information carried by the corresponding principal component, and the more significant the impact on the fund concentration degree. Therefore, when the cumulative variance ratio reaches a certain preset threshold (such as 85%), the present invention can reasonably infer that these principal components have fully explained most of the variance in the data.

[0142] Based on this judgment, the present invention can perform dimensionality reduction operations and retain only these principal components, thereby achieving effective dimensionality reduction of the data. This strategy not only helps to simplify the data analysis process, but also ensures that the data information retained by the present invention is sufficient to support subsequent decision-making and inference.

[0143] In an alternative embodiment, see Figure 6 , the constructing of the fund concentration model according to the principal component data comprises:

[0144] Step S3C4: Use the principal component data after dimensionality reduction as input features and the fund concentration degree as the target variable to construct a multivariate linear regression model of fund concentration degree.

[0145] Step S3C5: solving the model parameters of the multivariate linear regression model by the least square method, wherein the model parameters are grouped according to the dimensional conditions of time and enterprise code, and are aggregated on the dimension of a unique enterprise group.

[0146] The data after PCA dimensionality reduction is used as input features, and the fund concentration degree is used as the target variable.

[0147] A multivariate linear regression model was constructed and the model parameters (i.e., regression coefficients) were solved by the least squares method (OLS), see formula (3).

[0148] FCD=β0+β1×X1+β2×X2+···+β p ×X p +∈ (3)

[0150] Among them, FCD is defined as the degree of fund concentration, and X1, X2... represent a series of independent variables, such as: the amount of funds collected by the enterprise and the amount of funds not collected by the enterprise. These independent variables are grouped according to the dimension conditions of time (day) and enterprise code, and are aggregated on the dimension of enterprise group (only). β0, β1... represent regression coefficients, which are obtained by the influence coefficients of the features on the dependent variables after PCA data dimensionality reduction (see Section 3.3 for details). In addition, ∈ represents the error value, which reflects the difference between the actual observation value and the predicted value of the model independent variable and regression coefficient. Based on this error value, the present invention can perform a goodness of fit test, that is, when ∈ is close to 0, it means that the average value predicted by the model is closer to the average value of the actual observation.

[0151] Based on this method, the estimated degree of concentration of funds in member unit accounts under different degrees of influence of various conditions and factors is obtained.

[0152] In one embodiment of the method for calculating the concentration degree of enterprise account funds of the present invention, see Figure 7 The evaluation result of the current fund collection status of the enterprise is determined according to the calculation result of the fund concentration degree, including:

[0153] Step S13A: Determine the fund collection efficiency of multiple business systems according to the calculation result of the fund concentration degree.

[0154] After the calculation of the fund concentration model is completed, the fund collection efficiency in different business systems can be evaluated based on the results generated by the model. The fund collection efficiency reflects the management and collection of funds in each business module of the enterprise. By analyzing the collection of multiple systems (such as accounts receivable and payable, financing, settlement, etc.), the enterprise can clearly understand which systems have better centralized fund management effects and which systems may have problems of fund dispersion or inefficient management, which provides a basis for subsequent risk warnings and optimization suggestions.

[0155] Step S13B: Based on the analysis results of fund collection efficiency, for business systems with low collection rates, risk warnings are generated in combination with the fund flows and fluctuations of each system.

[0156] After evaluating the fund collection efficiency of each business system, the enterprise can generate risk warnings based on the systems with low collection rates. If the fund collection rate of a system is low, it indicates that its funds may not be effectively collected, and there is a risk of poor management or low fund utilization. Combining the fund flow and fund fluctuations of these systems, the enterprise can promptly identify potential risk points, such as insufficient fund liquidity and the risk of retention caused by fund dispersion. Through these warnings, the enterprise can take measures in advance to avoid possible fund management problems.

[0157] Step S13C: Generate optimization suggestions for enterprise fund collection based on the fund collection efficiency analysis results and the generated risk warning.

[0158] After generating a risk warning, the enterprise can also make targeted optimization suggestions based on the results of the collection efficiency analysis and risk warning. These optimization suggestions are aimed at improving the fund management efficiency of systems with low fund collection. For example, the enterprise may need to adjust the fund scheduling strategy, optimize the fund flow path, or strengthen the fund collection management of specific business systems. Through these optimization suggestions, the enterprise can better improve the utilization rate of funds and enhance the effect of centralized fund management, thereby improving overall operational efficiency and reducing fund risks.

[0159] In order to further illustrate the present solution, the present invention also provides an example of a method for calculating the concentration degree of funds in an enterprise account. Figure 8 As shown, the method of the present invention fully combines the advantages of big data and artificial intelligence technology, and provides comprehensive and accurate support for enterprise fund collection. At the business definition level of the model, the factors affecting enterprise fund collection are complex and diverse, including but not limited to bill processing, bank deposit management, accounts receivable and payable, etc. In order to fully capture these key factors affecting capital flow, the model constructed by the present invention conducts in-depth data mining and feature extraction, covering almost every business link and capital entrance of the enterprise financial system.

[0160] Based on the enterprise-level treasury data center, the present invention realizes the comprehensive collection and aggregation of enterprise and member unit data. Preliminary data cleaning and archiving are carried out through the data center to ensure the accuracy and reliability of the data. On this basis, the present invention further uses the principal component analysis (PCA) technology to reduce the dimensionality of the original high-dimensional data set. This step not only effectively removes noise and redundant information, but also retains the main features of the data, thereby improving the efficiency and accuracy of subsequent model training.

[0161] After the dimensionality reduction process is completed, the present invention inputs the optimized data into the multivariate linear regression model for training. Through this step, the present invention not only reduces the complexity of the model and improves the calculation efficiency, but also effectively avoids the occurrence of overfitting, thereby enhancing the generalization ability of the model. This significantly improves the accuracy of the fund concentration model, while also ensuring the comprehensiveness and integrity of the model.

[0162] In addition, the model not only helps companies identify potential risks in the fund collection process, but also provides decision support for management and optimizes fund allocation strategies. By regularly monitoring and analyzing the results of the model output, companies can promptly discover and solve problems in the fund collection process and ensure the efficiency and security of fund flow.

[0163] In summary, the present invention has achieved comprehensive capture and accurate analysis of the factors affecting the capital concentration of enterprises by comprehensively using big data and artificial intelligence technology. By optimizing the data processing and model training process, the accuracy and comprehensiveness of the capital concentration model have been improved, providing strong support for corporate capital management and decision-making.

[0164] The embodiment of the present invention also provides a device 002 for calculating the concentration degree of funds in an enterprise account, see Fig. 9 , including a data input module 21, a convergence degree calculation module 22 and a convergence degree evaluation module 23, wherein:

[0165] The data input module 21 is used to collect the capital concentration model pre-built from the capital flow data of multiple business systems in the enterprise, wherein the collection model is obtained by standardizing and reducing the dimension of the collected capital flow data through principal component analysis, and the capital flow data includes account balances, receivables and payables information, financing information, bill information and settlement information;

[0166] The concentration degree calculation module 22 is used to perform concentration analysis on the input capital flow data based on the capital concentration degree concentration model to obtain a calculation result of the capital concentration degree, wherein the calculation result represents the capital factors affecting the current capital concentration degree;

[0167] The concentration degree evaluation module 23 is used to determine the evaluation result of the current fund concentration status of the enterprise according to the calculation result of the fund concentration degree, and the evaluation result includes the fund concentration status of multiple business systems, risk warning and fund optimization suggestions.

[0168] According to any embodiment of the present invention, the process of constructing the fund concentration model includes:

[0169] Data collection module, used to collect fund flow data from multiple business systems through the enterprise-level treasury platform;

[0170] A data preprocessing module, used to preprocess the collected capital flow data and store it in the data center, wherein the preprocessing includes data cleaning, data conversion, data standardization, data integration and data compression;

[0171] The model building module is used to analyze the main factors affecting the concentration of funds in the fund flow data of the data center through principal component analysis to build the fund concentration model.

[0172] According to any embodiment of the present invention, the model building module includes:

[0173] An indicator determination unit is used to identify key factors affecting the degree of fund concentration through the fund flow data of the data center, and calculate fund indicator data according to the key factors, including uncollected funds within the attributable scope, collected funds within the attributable scope, uncollected funds within the full scope, collected funds within the full scope, uncollected fund structure analysis, total funds, total collectable funds, and total uncollected funds within the attributable scope;

[0174] A standardization processing unit is used to standardize the collected capital indicator data using principal component analysis to ensure that the weights of each feature in the analysis are consistent;

[0175] The data dimension reduction unit is used to convert the multi-dimensional capital indicator data into low-dimensional principal component data using principal component analysis, and to construct the capital concentration aggregation model based on the principal component data.

[0176] According to any embodiment of the present invention, the standardization processing unit is specifically used for

[0177] Calculate the mean and standard deviation of the collected capital indicator data to obtain the average level and dispersion of the capital indicator data respectively;

[0178] Each data point is converted into standardized data based on a standardization formula so that the standardized data will have distribution characteristics with a mean of zero and a standard deviation of one.

[0179] According to any embodiment of the present invention, the data dimension reduction unit is specifically used for

[0180] Calculate the covariance between data sets consisting of multiple fund indicator data to determine the volatility of the fund concentration in each indicator;

[0181] Use principal component analysis to determine the direction with the largest variance in the fund indicator data, and retain the principal components with variance ratios exceeding the preset value;

[0182] According to the amount of information contained in the principal components, the principal components with higher cumulative variance ratios are retained and the secondary components with smaller cumulative variance ratios are removed to achieve principal component dimensionality reduction.

[0183] According to any embodiment of the present invention, the data dimension reduction unit is also used for

[0184] Using the principal component data after dimensionality reduction as input features and the fund concentration degree as the target variable, a multivariate linear regression model of fund concentration degree is constructed;

[0185] The model parameters of the multivariate linear regression model are solved by the least square method, wherein the model parameters are grouped according to the dimension conditions of time and enterprise code, and are aggregated on the dimension of the only enterprise group.

[0186] According to any embodiment of the present invention, the aggregation degree evaluation module includes:

[0187] A collection efficiency determination unit, used to determine the fund collection efficiency of multiple business systems according to the calculation result of the fund concentration degree;

[0188] The risk warning unit is used to generate risk warnings based on the results of fund collection efficiency analysis for business systems with low collection rates, combined with the fund flows and fluctuations of each system;

[0189] The optimization suggestion unit is used to generate optimization suggestions for the enterprise's fund collection based on the fund collection efficiency analysis results and the generated risk warnings.

[0190] It should be noted that the method for calculating the concentration degree of enterprise account funds provided in the embodiment of the present invention can be used in the financial field, and can also be used in any technical field other than the financial field. The embodiment of the present invention does not limit the application field of the virtual reality-based bank training system and its training method.

[0191] As described in the above embodiment, since the principle of the device to solve the problem is similar to the method for calculating the concentration degree of enterprise account funds, the implementation of the device can refer to the implementation of the method for calculating the concentration degree of enterprise account funds, and the repeated parts will not be repeated.

[0192] Fig.10 A schematic diagram of the physical structure of an electronic device provided in an embodiment of the present invention, such as Fig.10 As shown, the electronic device 003 includes a processor 301, a memory 302 and a bus 303. The processor 301 and the memory 302 communicate with each other via the bus 303.

[0193] The processor 501 is used to call the program instructions in the memory 502 to execute the methods provided by the above-mentioned method embodiments.

[0194] An embodiment of the present invention further provides a computer-readable storage medium, wherein the computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the above-mentioned method for calculating the concentration degree of funds in an enterprise account is implemented.

[0195] An embodiment of the present invention further provides a computer program product, which includes a computer program. When the computer program is executed by a processor, the above-mentioned method for calculating the concentration degree of funds in an enterprise account is implemented.

[0196] Those skilled in the art will appreciate that embodiments of the present invention may be provided as methods, systems, or computer program products. Therefore, the present invention may take the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware. Moreover, the present invention may take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0197] The present invention is described with reference to flowcharts and / or block diagrams of methods, devices (systems), and computer program products according to embodiments of the present invention. It should be understood that each process and / or block in the flowchart and / or block diagram, as well as the combination of processes and / or blocks in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0198] These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.

[0199] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process. Figure 1 A process or multiple processes and / or boxes Figure 1 The steps for the functions specified in one or more boxes.

[0200] The specific embodiments described above further illustrate the objectives, technical solutions and beneficial effects of the present invention in detail. It should be understood that the above description is only a specific embodiment of the present invention and is not intended to limit the scope of protection of the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the scope of protection of the present invention.

Claims

1. A method for calculating the concentration degree of enterprise account funds, characterized in that: include: A fund concentration aggregation model pre-constructed from fund flow data of multiple business systems in an enterprise, wherein the aggregation model is obtained by standardizing and reducing the dimension of the collected fund flow data through principal component analysis, and the fund flow data includes account balances, accounts receivable and payable information, financing information, bill information and settlement information; Based on the fund concentration degree collection model, the input fund flow data is collected and analyzed to obtain a calculation result of the fund concentration degree, wherein the calculation result represents the fund factors that affect the current fund concentration degree; An evaluation result of the current fund collection status of the enterprise is determined according to the calculation result of the fund concentration degree, and the evaluation result includes the fund collection status of multiple business systems, risk warning and fund optimization suggestions.

2. The method for calculating the concentration degree of enterprise account funds according to claim 1, characterized in that: The construction process of the fund concentration aggregation model includes: Collect fund flow data from multiple business systems through an enterprise-level treasury platform; Preprocessing the collected capital flow data and storing it in the data center, wherein the preprocessing includes data cleaning, data conversion, data standardization, data integration and data compression; The principal component analysis is used to analyze the main factors affecting the fund concentration in the fund flow data of the data center to construct the fund concentration model.

3. The method for calculating the concentration degree of enterprise account funds according to claim 2, characterized in that: Determining the main factors affecting the fund concentration in the fund flow data of the data center through principal component analysis and constructing the fund concentration model includes: Identify the key factors that affect the degree of fund concentration through the fund flow data of the data center, and calculate the fund index data according to the key factors, including the uncollected funds within the attributable scope, the collected funds within the attributable scope, the uncollected funds within the full scope, the collected funds within the full scope, the structure analysis of the uncollected funds, the total amount of funds, the total amount of funds that can be collected, and the total amount of funds that cannot be collected within the attributable scope; The collected funding indicator data were standardized using principal component analysis to ensure that the weights of each feature in the analysis were consistent; The multi-dimensional capital index data is converted into low-dimensional principal component data using principal component analysis, and the capital concentration aggregation model is constructed based on the principal component data.

4. The method for calculating the concentration degree of enterprise account funds according to claim 3, characterized in that: The principal component analysis is used to standardize the collected capital indicator data to ensure that the weights of each feature in the analysis are consistent, including: Calculate the mean and standard deviation of the collected capital indicator data to obtain the average level and dispersion of the capital indicator data respectively; Each data point is converted into standardized data based on a standardization formula so that the standardized data will have distribution characteristics with a mean of zero and a standard deviation of one.

5. The method for calculating the concentration degree of enterprise account funds according to claim 3, characterized in that: The principal component analysis is used to convert multi-dimensional capital indicator data into low-dimensional principal component data, including: Calculate the covariance between data sets consisting of multiple fund indicator data to determine the volatility of the fund concentration in each indicator; Use principal component analysis to determine the direction with the largest variance in the fund indicator data, and retain the principal components with variance ratios exceeding the preset value; According to the amount of information contained in the principal components, the principal components with higher cumulative variance ratios are retained and the secondary components with smaller cumulative variance ratios are removed to achieve principal component dimensionality reduction.

6. The method for calculating the concentration degree of enterprise account funds according to claim 3, characterized in that: The constructing the fund concentration aggregation model according to the principal component data includes: Using the principal component data after dimensionality reduction as input features and the fund concentration degree as the target variable, a multivariate linear regression model of fund concentration degree is constructed; The model parameters of the multivariate linear regression model are solved by the least square method, wherein the model parameters are grouped according to the dimension conditions of time and enterprise code, and are aggregated on the dimension of the only enterprise group.

7. The method for calculating the concentration degree of enterprise account funds according to claim 1, characterized in that: The evaluation result of the current fund collection status of the enterprise is determined according to the calculation result of the fund concentration degree, including: Determining the fund collection efficiency of multiple business systems according to the calculation result of the fund concentration degree; Based on the results of fund collection efficiency analysis, for business systems with low collection rates, risk warnings are generated based on the fund flow and fluctuations of each system; Generate optimization suggestions for enterprise fund collection based on the fund collection efficiency analysis results and generated risk warnings.

8. A device for calculating the concentration degree of enterprise account funds, characterized in that: include: A data input module is used to pre-construct a fund concentration aggregation model for fund flow data of multiple business systems in an enterprise, wherein the aggregation model is obtained by standardizing and reducing the dimension of the collected fund flow data through principal component analysis, and the fund flow data includes account balances, receivables and payables information, financing information, bill information and settlement information; A concentration degree calculation module, used for performing concentration analysis on the input capital flow data based on the capital concentration degree concentration model to obtain a calculation result of the capital concentration degree, wherein the calculation result represents the capital factors affecting the current capital concentration degree; The concentration degree evaluation module is used to determine the evaluation result of the enterprise's current fund collection status according to the calculation result of the fund concentration degree, and the evaluation result includes the fund collection status of multiple business systems, risk warning and fund optimization suggestions.

9. A computer device comprising a memory, a processor and a computer program stored in the memory and executable on the processor, characterized in that: When the processor executes the computer program, the method for calculating the concentration degree of enterprise account funds described in any one of claims 1 to 7 is implemented.

10. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the method for calculating the concentration degree of funds in an enterprise account described in any one of claims 1 to 7 is implemented.

11. A computer program product, characterized in that The computer program product includes a computer program, and when the computer program is executed by a processor, it implements the method for calculating the concentration degree of funds in an enterprise account as described in any one of claims 1 to 7.

Images