Enterprise screening method, device and equipment and storage medium

Abstract

The invention discloses an enterprise screening method and device, equipment and a storage medium, and relates to the technical field of computers, and the method comprises the steps: obtaining financial data of a to-be-invested enterprise, and carrying out the preprocessing; analyzing financial statements and financial original variables of the preprocessed data, determining first and second model input and output variable systems, and constructing first and second enterprise screening models. And determining a relative importance index value of each variable based on the financial data, and constructing a third enterprise screening model by using the determined target variable. And performing efficiency evaluation on the preprocessed data by using the three models to obtain corresponding results. And performing model screening based on the result, determining a target enterprise screening model, screening the to-be-invested enterprises by using the corresponding target evaluation result and a preset efficiency threshold, and finally obtaining a screening result. Therefore, through multi-model construction and screening, the accuracy and reliability of enterprise screening are improved, and investment reference is provided for users.

Description

Technical Field

[0001] The present invention relates to the field of computer technology, and in particular to an enterprise screening method, device, equipment and storage medium. Background Art

[0002] In today's economic environment, as an important force for economic development, the investment value of small and medium-sized enterprises is increasingly prominent. However, investment in small and medium-sized enterprises faces many challenges. On the one hand, there is a serious asymmetry in market information, and it is difficult for investors to fully understand the real operating conditions, financial conditions and development potential of enterprises. On the other hand, the complexity of the market makes it extremely difficult to evaluate the performance of enterprises. Traditional evaluation methods are often limited to a single financial indicator or subjective judgment, and cannot accurately reflect the comprehensive strength and future development prospects of enterprises. In addition, small and medium-sized enterprises in different industries have their own characteristics, which increases the difficulty of evaluation.

[0003] Therefore, how to improve the efficiency and reliability of screening companies that provide investment services is an urgent problem that needs to be solved. Summary of the invention

[0004] In view of this, the purpose of the present invention is to provide a method, device, equipment and storage medium for enterprise screening, which can improve the efficiency and reliability of screening enterprises with investment services. The specific scheme is as follows:

[0005] In a first aspect, the present application provides a method for screening an enterprise, comprising:

[0006] Acquire financial data corresponding to the enterprises to be invested in several industries, and pre-process the financial data based on preset data pre-processing rules to obtain corresponding pre-processed data;

[0007] By respectively analyzing the financial statement data and the original financial variables in the pre-processed data corresponding to the enterprise to be invested, a first model input-output variable system and a second model input-output variable system are determined, and a first enterprise screening model and a second enterprise screening model are respectively constructed based on a preset data envelopment analysis technology and the first model input-output variable system and the second model input-output variable system;

[0008] Determine the relative importance index values ​​corresponding to the variables in the first model input-output variable system and the second model input-output variable system based on the financial data, and construct a third enterprise screening model based on the preset data envelopment analysis technology and the target variables determined by using the relative importance index values;

[0009] Respectively performing efficiency evaluation on the preprocessed data based on the first enterprise screening model, the second enterprise screening model, and the third enterprise screening model to obtain corresponding first evaluation results, second evaluation results, and third evaluation results;

[0010] Model analysis and screening are performed based on the first evaluation result, the second evaluation result and the third evaluation result, and the target evaluation results corresponding to the determined target enterprise screening model and the preset efficiency threshold are used to screen the enterprises to be invested in order to obtain corresponding screening results.

[0011] Optionally, preprocessing the financial data based on a preset data preprocessing rule to obtain corresponding preprocessed data includes:

[0012] Eliminate data that meets a first preset data missing condition in the financial data, and perform missing value interpolation on data that meets a second preset data missing condition in the financial data, so as to complete a first data processing operation and obtain corresponding first data;

[0013] Based on the average values ​​and standard deviations respectively corresponding to the financial variables in the first data, it is judged whether there is target data in the first data that meets the preset extreme value condition, so as to obtain a corresponding first judgment result;

[0014] If the first judgment result indicates that the target data exists, the target data is eliminated to complete the second data processing operation and obtain pre-processed data.

[0015] Optionally, the determining of the first model input-output variable system and the second model input-output variable system by respectively analyzing the financial statement data and the original financial variables in the pre-processed data corresponding to the enterprise to be invested in, comprises:

[0016] By analyzing the financial performance indicators in the financial statement data in the pre-processed data, a first model input and output variable system is constructed;

[0017] The input and output variable system of the second model is constructed by analyzing the original financial variables in the preprocessed data.

[0018] Optionally, the performing model analysis and screening based on the first evaluation result, the second evaluation result, and the third evaluation result includes:

[0019] Determine the mean of the efficiency values ​​of the first enterprise screening model, the second enterprise screening model and the third enterprise screening model for evaluating the enterprises to be invested in each industry based on the first evaluation result, the second evaluation result and the third evaluation result, and perform model analysis based on the obtained mean efficiency values ​​of each industry to obtain corresponding model analysis results;

[0020] A target enterprise screening model is screened out from the first enterprise screening model, the second enterprise screening model and the third enterprise screening model based on the model analysis result.

[0021] Optionally, the use of the target evaluation results corresponding to the determined target enterprise screening model and a preset efficiency threshold to screen the enterprises to be invested in, so as to obtain corresponding screening results, includes:

[0022] Obtaining target evaluation results corresponding to the determined target enterprise screening model;

[0023] Determine an average efficiency value corresponding to the enterprise to be invested based on the target evaluation result, and judge whether the average efficiency value is greater than a preset efficiency threshold to obtain a corresponding second judgment result;

[0024] Based on the second judgment result, the enterprises to be invested are screened to obtain corresponding screening results.

[0025] Optionally, determining, based on the financial data, relative importance index values ​​corresponding to the variables in the first model input-output variable system and the second model input-output variable system, respectively, includes:

[0026] Determine, based on the financial data, the relative weight values ​​of the variables in the first model input-output variable system and the second model input-output variable system in each time period and in each industry;

[0027] The average value of the relative weight value of each variable is calculated based on the corresponding number of time periods to obtain the relative importance index value of each variable in each industry.

[0028] Optionally, the step of constructing a third enterprise screening model based on the preset data envelopment analysis technology and the target variable determined by using each of the relative importance index values ​​includes:

[0029] respectively judging whether the relative importance index value of each of the variables in each of the industries is not less than a preset importance threshold, so as to obtain a corresponding third judgment result;

[0030] Counting the target number of the third judgment result corresponding to each of the variables and representing a value not less than the preset importance threshold, and judging whether each of the target numbers is not less than the target threshold, so as to obtain a corresponding fourth judgment result;

[0031] If the fourth judgment result indicates that it is not less than the target threshold, the corresponding variable is determined to be the target variable, and a third enterprise screening model is constructed based on the preset data envelopment analysis technology and the target variable.

[0032] In a second aspect, the present application provides an enterprise screening device, comprising:

[0033] A data preprocessing module, used to obtain financial data corresponding to the enterprises to be invested in several industries, and preprocess the financial data based on preset data preprocessing rules to obtain corresponding preprocessed data;

[0034] A first model building module is used to determine a first model input-output variable system and a second model input-output variable system by respectively analyzing the financial statement data and the original financial variables in the pre-processed data corresponding to the enterprise to be invested, and to respectively build a first enterprise screening model and a second enterprise screening model based on a preset data envelopment analysis technology and the first model input-output variable system and the second model input-output variable system;

[0035] A second model building module is used to determine the relative importance index values ​​corresponding to the variables in the first model input-output variable system and the second model input-output variable system based on the financial data, and to build a third enterprise screening model based on the preset data envelopment analysis technology and the target variables determined by using the relative importance index values;

[0036] an efficiency evaluation module, used to perform efficiency evaluation on the preprocessed data based on the first enterprise screening model, the second enterprise screening model, and the third enterprise screening model, respectively, to obtain corresponding first evaluation results, second evaluation results, and third evaluation results;

[0037] An enterprise screening module is used to perform model analysis and screening based on the first evaluation result, the second evaluation result and the third evaluation result, and to screen the enterprises to be invested using the target evaluation results corresponding to the determined target enterprise screening model and a preset efficiency threshold to obtain corresponding screening results.

[0038] In a third aspect, the present application provides an electronic device, including:

[0039] Memory, used to store computer programs;

[0040] The processor is used to execute the computer program to implement the aforementioned enterprise screening method.

[0041] In a fourth aspect, the present application provides a computer-readable storage medium for storing a computer program; wherein, when the computer program is executed by a processor, the aforementioned enterprise screening method is implemented.

[0042] In the present application, financial data corresponding to enterprises to be invested in several industries are obtained, and the financial data are preprocessed based on preset data preprocessing rules to obtain corresponding preprocessed data; the financial statement data and the original financial variables in the preprocessed data corresponding to the enterprises to be invested are respectively analyzed to determine the first model input-output variable system and the second model input-output variable system, and the first enterprise screening model and the second enterprise screening model are respectively constructed based on the preset data envelopment analysis technology and the first model input-output variable system and the second model input-output variable system; based on the financial data, the first model input-output variable system and the second model input-output variable system are determined. The relative importance index values ​​corresponding to the variables of the data are obtained, and a third enterprise screening model is constructed based on the preset data envelopment analysis technology and the target variables determined by the relative importance index values; the efficiency of the pre-processed data is evaluated based on the first enterprise screening model, the second enterprise screening model and the third enterprise screening model to obtain the corresponding first evaluation result, the second evaluation result and the third evaluation result; model analysis and screening are performed based on the first evaluation result, the second evaluation result and the third evaluation result, and the target evaluation result corresponding to the determined target enterprise screening model and the preset efficiency threshold are used to screen the enterprises to be invested to obtain the corresponding screening results. As can be seen from the above, in this application, the financial data of the enterprise to be invested is first obtained and pre-processed. The pre-processed data is analyzed to determine the input and output variable system of the first and second models, and the first and second enterprise screening models are constructed. The relative importance index values ​​of the variables are determined based on the financial data, and the third enterprise screening model is constructed. The efficiency of the data is evaluated using three models, and the model analysis and screening are performed based on the results to determine the target enterprise screening model. Finally, the target evaluation results and the preset efficiency threshold are used to screen the enterprises to be invested, and the screening results are obtained. In this way, this application can effectively improve the accuracy and reliability of enterprise screening for users to use as investment reference. BRIEF DESCRIPTION OF THE DRAWINGS

[0043] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on the provided drawings without paying creative work.

[0044] Figure 1 A flow chart of an enterprise screening method disclosed in this application;

[0045] Figure 2 A line chart of the relative importance index values ​​of input and output variables in various industries in a first enterprise screening model disclosed in the present application;

[0046] Figure 3 A line chart of the relative importance index values ​​of input and output variables in various industries in a second enterprise screening model disclosed in the present application;

[0047] Figure 4 A mean bar chart of the efficiency values ​​of the enterprises to be invested in various industries evaluated based on the enterprise screening models disclosed in this application;

[0048] Figure 5 A bar chart showing the proportion of the efficiency values ​​of the enterprises to be invested in various industries with an average value of 1 based on the enterprise screening models disclosed in this application;

[0049] Figure 6 This is a schematic diagram of the structure of an enterprise screening device disclosed in this application;

[0050] Figure 7 This is a structural diagram of an enterprise screening equipment disclosed in this application. DETAILED DESCRIPTION

[0051] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0052] In today's economic environment, small and medium-sized enterprises, as an important force for economic development, have become increasingly valuable for investment. However, investment in small and medium-sized enterprises faces many challenges. On the one hand, there is a serious asymmetry in market information, and it is difficult for investors to fully understand the company's true operating conditions, financial conditions, and development potential. On the other hand, the complexity of the market makes it extremely difficult to evaluate the performance of a company. Traditional evaluation methods are often limited to a single financial indicator or subjective judgment, and cannot accurately reflect the company's comprehensive strength and future development prospects. In addition, small and medium-sized enterprises in different industries have their own characteristics, which increases the difficulty of evaluation. To this end, this application provides a business screening solution to improve the efficiency and reliability of screening companies with investment services.

[0053] See also Figure 1 As shown, the embodiment of the present invention discloses a method for screening enterprises, including:

[0054] S11. Obtain financial data corresponding to the enterprises to be invested in several industries, and preprocess the financial data based on preset data preprocessing rules to obtain corresponding preprocessed data.

[0055] In this embodiment, it is first necessary to obtain financial data corresponding to the enterprises to be invested in several industries. The enterprises to be invested can be small and medium-sized enterprises to be invested. There are a large number of small and medium-sized enterprises, which are widely distributed, covering various industries and fields, and have innovative vitality and can quickly adapt to market changes. Therefore, screening small and medium-sized enterprises to be invested helps investors find potential investment targets. Among them, financial data is the basis for screening and evaluating enterprises to be invested. Financial data includes but is not limited to data in various financial statements such as the balance sheet, income statement, cash flow statement, etc. of the enterprise, as well as market data, etc. Further, in order to ensure the integrity and reliability of the financial data, it is necessary to pre-process the financial data based on the preset data pre-processing rules to obtain the corresponding pre-processed data.

[0056] It should be understood that, in this embodiment, the preprocessing process is mainly divided into two main data processing operations. First, the first data processing operation is performed. In this process, the data in the financial data that meets the first preset data missing condition is eliminated. The first preset data missing condition can be set as a situation where a large amount of financial data is missing. For example, if a large amount of financial data of a certain enterprise is missing, or the financial data is missing for a relatively long continuous period of time, then the financial data corresponding to the enterprise to be invested will be deleted. In addition, financial data that is missing due to suspension of trading during a specific period, or financial data that is missing due to a delisting risk warning, can be deleted. In this way, the problem of inaccurate subsequent analysis caused by serious lack of financial data can be avoided.

[0057] At the same time, the missing value interpolation is performed on the data in the financial data that meets the second preset data missing condition to complete the first data processing operation. The second preset data missing condition can be set as a situation where a small amount of data is missing. Retaining these financial data with a small amount of missing data can make up for the incompleteness of the data to a certain extent and improve the availability of the data.

[0058] In this embodiment, after the first data processing operation, the corresponding first data is obtained. Next, in order to avoid the adverse effects of extreme values ​​on subsequent enterprise screening and evaluation, further, based on the average values ​​and standard deviations corresponding to each financial variable in the first data, it is judged whether there is target data that meets the preset extreme value conditions in the first data to obtain the corresponding first judgment result. If the first judgment result indicates that the target data exists, the target data is eliminated to complete the second data processing operation and obtain the preprocessed data. Specifically, first calculate the average value and standard deviation of each financial variable in the first data, and then judge whether there is financial data in the first data whose deviation from the average value exceeds the preset multiple standard deviation, so as to obtain the corresponding first judgment result. If there is financial data whose deviation from the average value exceeds the preset multiple standard deviation, the financial data is determined as the target data and the target data is eliminated. For example, the average value of the financial variable is set to , the standard deviation is set to , then delete or less than financial data.

[0059] S12. Determine the first model input-output variable system and the second model input-output variable system by respectively analyzing the financial statement data and the original financial variables in the preprocessed data corresponding to the enterprise to be invested, and respectively construct the first enterprise screening model and the second enterprise screening model based on the preset data envelopment analysis technology and the first model input-output variable system and the second model input-output variable system.

[0060] In this embodiment, after the preprocessing of the financial data is completed, the first model input and output variable system is constructed by analyzing the financial performance indicators in the financial statement data in the preprocessed data;

[0061] The financial original variables in the pre-processed data are analyzed to construct the second model input and output variable system. Then, based on the preset data envelopment analysis technology and the constructed two variable systems, the first enterprise screening model and the second enterprise screening model are respectively constructed.

[0062] It can be understood that, in this embodiment, the input-output variable system of the first model is constructed by analyzing the financial performance indicators in the financial statement data in the preprocessed data. The financial statement data contains relatively comprehensive operating information of the enterprise to be invested. By analyzing the financial performance indicators commonly used in the accounting industry in the operating information, financial performance indicators that can objectively reflect the basic characteristics of the industry in which the enterprise to be invested is located can be selected. It should be noted that in the selection process, it is necessary to ensure that strong linear relationships between the various financial performance indicators are avoided as much as possible, and the availability of financial performance indicators must also be considered. For example, as shown in Table 2 below, from the quarterly financial statements, the debt-to-asset ratio, long-term liabilities / shareholders' equity, current assets / total assets, etc. can be selected as input variables, and the earnings per share growth rate, operating income growth rate, fixed asset turnover rate, etc. can be selected as output variables, and then the first model input-output variable system is constructed based on the above input variables and output variables.

[0063] It is understandable that financial performance indicators such as the debt-to-asset ratio reflect the capital structure and debt repayment ability of the company to be invested. Taking them as input variables can measure the investment situation of the company to be invested; while financial performance indicators such as the earnings per share growth rate reflect the profitability and development potential of the company to be invested. Taking them as output variables can measure the output performance of the company to be invested.

[0064] Next, in this embodiment, the input and output variable system of the second model is constructed by analyzing the original financial variables in the preprocessed data. The original financial variables can reflect the operating conditions of the enterprise to be invested. Specifically, the operating income and net profit can be set as output variables, and the accounts receivable, inventory, total assets, and total liabilities of the enterprise to be invested can be set as input variables. It can be understood that the operating income and net profit directly reflect the operating results of the enterprise to be invested, while the accounts receivable and inventory reflect the resource occupation of the enterprise to be invested in the operation process, and the total assets and total liabilities reflect the scale and debt repayment pressure of the enterprise to be invested. Through the analysis of these original financial variables, the input and output variable system of the second model is constructed to provide a basis for the construction of the second enterprise screening model.

[0065] In this embodiment, after the input and output variable systems of the two models are determined, the first enterprise screening model and the second enterprise screening model are respectively constructed based on the preset data envelopment analysis technology and the first model input and output variable system and the second model input and output variable system.

[0066] It should be noted that DEA (Data Envelopment Analysis) is a non-parametric mathematical programming method for estimating the production frontier. In this embodiment, the preset data envelopment analysis technology is the modified CCR (Charnes-Cooper-Rhodes, DEA technology created by Charnes, Cooper and Rhodes) algorithm of this embodiment.

[0067] Specifically, regarding the modified CCR algorithm, the enterprise screening model constructed in this embodiment is a CCR model. The CCR model reflects the efficiency of the target DMU (i.e., Decision Making Unit) by the ratio between the comprehensive output and the comprehensive input. Its characteristic is that variability weights are given to the model input and output variables. The decision making unit is the enterprise to be invested in this embodiment. The efficiency value of the target DMU can be obtained by solving the following fractional programming problem:

[0068] ;

[0069] In the formula, and Respectively The input vector and output vector of the target DMU ( );and and are the weight vectors corresponding to the input and output vectors respectively; T represents the transpose of the weight vector; represents the transpose of the e matrix; Indicates that the input weights cannot all be 0; .. means under the condition of .., Indicates the maximum value; It represents the ratio of the output vector of the evaluated decision-making unit to the input vector; Represents the ratio of the output vector of other decision-making units to the input vector. Here, the target DMU is the observed One of the DMUs, its input vector and output vector are recorded as and In addition, it should be explained that the input-output vector refers to the target DMU in the formula, which represents the input and output of the target DMU, ​​while the input-output variables are the indicators selected when building the CCR model. For the convenience of calculation, the Charners-Cooper transformation is used to transform the fractional programming problem into an equivalent linear programming problem, as shown below, that is, the input-oriented CCR model:

[0070] ;

[0071] In the formula, is the weight of the input vector, Transpose it; The input vector of the decision-making unit being evaluated is multiplied by its weight; The input vector of other decision-making units is multiplied by their weights; is the weight of the output vector, Transpose it; The output vector of the decision-making unit being evaluated is multiplied by its weight; is the output vector of other decision units multiplied by their weights. The dual problem is:

[0072] ;

[0073] In the dual programming, λ j Represents the linear combination coefficient of other decision-making units; parameter θ is the efficiency value (which can be measured using the ratio of output to input), which ranges from 0 to 1.

[0074] From a formal point of view, the input-oriented CCR model examines the possibility that the input of the target DMU can be reduced under the condition that the output does not decrease. Obviously, the optimal value of the problem is not greater than 1. If the optimal value is 1, the target DMU is called DEA weakly effective, otherwise it is non-DEA weakly effective. When the DMU is DEA weakly effective, under the condition of given output, the input can no longer be reduced, thus showing effectiveness in the Pareto sense. The traditional CCR model requires that the values ​​of input and output indicators are positive. In this embodiment, the values ​​of the input and output variables may be zero or negative, such as growth rate, return on assets, etc. If the values ​​of the input indicators are all negative, the constraints in the output-oriented performance evaluation model It is meaningless and the model needs to be modified. In view of the negative index, the constraint conditions of the CCR model are modified. The modified CCR algorithm model is as follows:

[0075] ;

[0076] In the above modified CCR algorithm model, if the input variable is greater than zero, the average efficiency value obtained by the modified CCR algorithm model is equal to the average efficiency value obtained by the traditional CCR model. If the input index is less than or equal to zero, the average efficiency value obtained by the modified CCR algorithm model is zero.

[0077] In addition, in this embodiment, in order to define the production technology status from an empirical level, a production possibility set can be constructed based on a series of axiom systems. Assume that the observed DMU data is:

[0078] ;

[0079] in, Respectively represent The input vector and output vector of a DMU. From an economic point of view, the input vector generally represents the input of production factors, while the output vector represents the corresponding number of different products.

[0080] If the input and output of a production system are and , then the corresponding production possibility set The mathematical representation of is:

[0081] ;

[0082] That is, the production possibility set represents a technical possibility, and does not mean that there is an effective correspondence between input and output. The following axioms may be satisfied:

[0083] Convexity axiom: requires production of possible sets is a convex set, or, in other words, for any , and any , both have:

[0084] ;

[0085] Trivial axiom: The observed DMU must belong to the production possibility set , or in other words, what has happened must be possible, that is:

[0086] ;

[0087] Inefficiency axiom: that is, the axiom of free disposal, which allows for the possibility of inefficiency in reality. , then when Sometimes, there are That is, if a given input-output combination is technically feasible, then it is also technically feasible to have more input and less output. The invalidity axiom can also be divided more specifically into the invalid input or invalid output axiom.

[0088] Cone axiom: also known as the additivity axiom, that is, the invariance of returns to scale. , then when Sometimes, there are .

[0089] Contraction axiom: This assumption corresponds to the non-increasing returns to scale. , then when Sometimes, there are .

[0090] Expansion axiom: This assumption corresponds to the non-decreasing returns to scale. , then when Sometimes, there are .

[0091] Minimalism Axiom: Production Possibility Set It is the smallest one that satisfies the given axioms, or in other words, it is the intersection of all possible production sets that satisfy the given axioms. It is precisely because of the minimality that the possible production set has unique certainty.

[0092] Given a DMU and constructing a reasonable axiom system, a unique DEA production possibility set can be determined, thereby defining the production technology status. For example, if it is assumed that the production possibility set satisfies convexity, triviality, inefficiency, cone and minimality, the corresponding production technology can be expressed as the following CCR production possibility set: :

[0093] ;

[0094] Under the condition of as little input as possible and as much output as possible, the (weak) Pareto efficient points in the CCR production possibility set can be similarly defined. If a point is a Pareto point, there are no other points in the CCR production possibility set, with less input and more output. All (weak) Pareto points in the CCR production possibility set constitute the (weak) efficient frontier of the CCR production possibility set, which is an important reference for subsequent efficiency evaluation. In addition, it should be noted that the model input and output variable system and the construction of the model refer to this axiom system.

[0095] S13. Determine the relative importance index values ​​corresponding to the variables in the first model input-output variable system and the second model input-output variable system based on the financial data, and construct a third enterprise screening model based on the preset data envelopment analysis technology and the target variables determined by the relative importance index values.

[0096] In this embodiment, in order to improve the accuracy and reliability of enterprise screening, it is first necessary to determine the relative weight values ​​of each variable in the first input-output variable system and the second input-output variable system in each time period and in each industry based on the financial data; then, based on the corresponding number of time periods (represented by N), calculate the average value of the relative weight values ​​of each variable to obtain the relative importance index value of each variable in each industry.

[0097] Specifically, taking the first model input and output variable system as an example, when solving the first enterprise screening model, if the variable If the optimal weight is not zero, then The relative weight value is ; If the variable The optimal weight of is zero, then The relative weight of Among them, the variable The calculation formula for the relative weight value in industry j at time t is as follows:

[0098] ;

[0099] In the formula, k is the serial number of the enterprise to be invested. Calculate the average value of the relative weight value of N time periods. Among them, the variable The calculation formula for the average relative weight value in N time periods is as follows:

[0100] ;

[0101] In the formula, is the variable in industry j The average value of the relative weight value, the relative weight value range is The closer its value is to 1, the greater the role the variable plays in the subsequent calculation of the average efficiency value.

[0102] In the process of calculating the average efficiency value of the enterprise to be invested by using the enterprise screening model, when the sample size remains unchanged, the increase of variables will cause the efficiency score to lose the ability to distinguish good and bad enterprises. Since the sample size of the present invention is fixed, the discrimination degree of the model for small and medium-sized enterprises can be increased by reducing the number of input or output variables. For this reason, the present invention will select more suitable variables in the first model input and output variable system and the second model input and output variable system according to the size of the variable relative importance index value, and use the selected variables to construct a new enterprise screening model, that is, the third enterprise screening model.

[0103] Specifically, after obtaining the relative importance index values ​​corresponding to each variable in the first model input-output variable system and the second model input-output variable system, respectively, determine whether the relative importance index values ​​of each variable in each industry are not less than the preset importance threshold to obtain the corresponding third judgment result; count the target number of the third judgment results corresponding to each quantity that is not less than the preset importance threshold, and determine whether each target number is not less than the target threshold to obtain the corresponding fourth judgment result; if the fourth judgment result shows that it is not less than the target threshold, determine the corresponding variable as the target variable, and construct a third enterprise screening model based on the preset data envelopment analysis technology and the target variable.

[0104] S14. Evaluate the efficiency of the preprocessed data based on the first enterprise screening model, the second enterprise screening model, and the third enterprise screening model to obtain corresponding first evaluation results, second evaluation results, and third evaluation results.

[0105] In this embodiment, the preprocessed data is input into the first enterprise screening model, the second enterprise screening model and the third enterprise screening model. The enterprise screening model will evaluate the efficiency of the enterprises to be invested in various industries based on the input preprocessed data, thereby obtaining the corresponding first evaluation results, second evaluation results and third evaluation results.

[0106] S15. Perform model analysis and screening based on the first evaluation result, the second evaluation result and the third evaluation result, and use the target evaluation result corresponding to the determined target enterprise screening model and the preset efficiency threshold to screen the enterprises to be invested in order to obtain corresponding screening results.

[0107] In this embodiment, based on the first evaluation result, the second evaluation result and the third evaluation result, the average of the efficiency values ​​of the first enterprise screening model, the second enterprise screening model and the third enterprise screening model for evaluating the enterprises to be invested in each industry is determined, and model analysis is performed based on the obtained efficiency averages of each industry to obtain corresponding model analysis results; then, based on the model analysis results, the target enterprise screening model is screened out from the first enterprise screening model, the second enterprise screening model and the third enterprise screening model.

[0108] Based on the analysis of the mean efficiency of each industry and the proportion of the mean efficiency of each industry being 1 of the three enterprise screening models, it can be explained that although the mean efficiency of each industry based on the first enterprise screening model is relatively high, the proportion of the mean efficiency of each industry being 1 is too high, resulting in insufficient differentiation of the enterprises to be invested; the mean efficiency of each industry and the proportion of the mean efficiency of each industry being 1 of the second enterprise screening model are both too low, which will also result in insufficient differentiation of the enterprises to be invested; and the mean efficiency of each industry of the third enterprise screening model is relatively moderate. Therefore, from the perspective of the mean efficiency of each industry based on each model, the third enterprise screening model is most suitable for the selection of enterprises to be invested, that is, the third enterprise screening model can be determined as the target enterprise screening model.

[0109] After determining the target enterprise screening model, obtain the target evaluation result corresponding to the determined target enterprise screening model; determine the average efficiency value corresponding to the enterprise to be invested based on the target evaluation result, and judge whether the average efficiency value is greater than the preset efficiency threshold to obtain the corresponding second judgment result; screen the enterprise to be invested based on the second judgment result to obtain the corresponding screening result. Among them, the preset efficiency threshold is a standard value set according to actual investment needs and experience, which is used to judge whether the enterprise to be invested has investment value. If the average efficiency value of the enterprise to be invested is greater than the preset efficiency threshold, it means that the enterprise to be invested performs well under the target enterprise screening model, has high efficiency and potential, and may be a worthy investment object; if the average efficiency value is less than or equal to the preset efficiency threshold, it means that the enterprise to be invested performs poorly under the target enterprise screening model and may not have investment value. As can be seen from the above, in this application, the financial data of the enterprise to be invested is first obtained and preprocessed. Analyze the preprocessed data, determine the input and output variable system of the first and second models, and construct the first and second enterprise screening models. Then determine the relative importance index value of the variable based on the financial data, and construct the third enterprise screening model. Use three models to evaluate the efficiency of the data, analyze and screen the models based on the results, and determine the target enterprise screening model. Finally, use the target evaluation results and the preset efficiency threshold to screen the investment enterprises and obtain the screening results. In this way, this application can effectively improve the accuracy and reliability of enterprise screening for users to make investment references.

[0110] Combine the following Figure 2-5 The line graph and bar graph disclosed in the text specifically illustrate the technical solutions of the embodiments of the present application.

[0111] In a specific implementation, the quarterly financial data of all non-financial and insurance small and medium-sized enterprises in a certain place from January 2010 to September 2018 are used as research samples for enterprise screening. In addition, it should be noted that in this embodiment, the selection of research samples does not limit the number of industries, types of industries, and the number of enterprises to be invested.

[0112] In this embodiment, there are multiple industries and multiple enterprises to be invested in the research sample. According to the classification of the Small and Medium Enterprises Industry Classification Guide (revised in 2012), 8 industries such as mining and electric power are selected as industries in the research sample. The number of enterprises to be invested in each industry is shown in Table 1 below, among which there are 45 enterprises to be invested in the mining industry, 74 enterprises to be invested in the production and supply of electricity, heat, gas and water (referred to as electricity), 87 enterprises to be invested in the real estate industry, 32 enterprises to be invested in the construction industry, 60 enterprises to be invested in the transportation, warehousing and postal industry (referred to as transportation), 100 enterprises to be invested in the wholesale and retail industry (referred to as wholesale and retail), 56 enterprises to be invested in the information software and information technology service industry (referred to as information service), and 679 enterprises to be invested in the manufacturing industry.

[0113] Table 1

[0114]

[0115] In this embodiment, the financial data in the research sample needs to be preprocessed. First, it is necessary to eliminate incomplete data in the research sample. Specifically, delete the financial data with serious data missing between 2010 and 2019, delete the financial data with serious data missing due to suspension during this period, and delete the financial data corresponding to the companies to be invested that have been issued delisting risk warnings during this period. However, financial data with a small amount of missing data can be retained, and the missing financial data can be supplemented by vector autoregression. Then, in order to make the subsequent calculation results more realistic and reliable, it is necessary to delete the financial data whose mean value deviation corresponding to the financial variables exceeds or is less than 3 times the standard deviation. Finally, after the preprocessing operation of the research sample, there are a total of 1,133 companies to be invested in the research sample at this time.

[0116] It can be understood that in this embodiment, the financial data of 1133 companies to be invested need to be input into the constructed first enterprise screening model and the second enterprise screening model to obtain the efficiency average of each industry output by the first enterprise screening model and the second enterprise screening model. As shown in Table 2 below, the input variables of the first enterprise screening model are debt-to-asset ratio (P1), long-term debt / shareholder equity (P2), current assets / total assets (P3), price-to-book ratio (P4) and management expense ratio (P5), and the output variables are earnings per share growth rate (P6), operating income growth rate (P7), fixed asset turnover rate (P8), current ratio (P9), quick ratio (P10), and operating income growth rate (P11). 10 ), inventory turnover rate (P 11 ), accounts receivable turnover ratio (P 12 ), earnings per share (P 13 ), net assets per share (P 14 ), return on net assets (P 15), return on assets (P 16 ) and return on invested capital (P 17 ), so the input and output variable system of the first model is as follows:

[0117] ;

[0118] In this embodiment, the output variables of the second enterprise screening model are operating income (Q1) and net profit (Q2), and the input variables are accounts receivable (Q3), inventory (Q4), total assets (Q5), and total liabilities (Q6). Therefore, the input and output variable system of the second model is as follows:

[0119] ;

[0120] Table 2

[0121]

[0122] In order to determine the relative importance of each variable in the enterprise screening model, this embodiment will calculate the relative weight values ​​of the two enterprise screening model variables for 35 periods in 8 industries, and then further calculate the average value of the relative weight values ​​based on the relative weight values, that is, the relative importance index value. Among them, the average value calculation formula of the relative weight values ​​of the enterprise screening model variables is as follows:

[0123] ;

[0124] In this embodiment, Figure 2 and Figure 3 To screen the relative importance index values ​​of input and output variables in the model for enterprises in various industries. Figure 2 The relative importance index values ​​of the input and output variables in each industry in the first enterprise screening model are Figure 2 It can be seen that in the first enterprise screening model, the relative importance index value of the variable of management expense ratio is relatively high, that is, the relative importance is relatively high, while the relative importance index value of the variable of current ratio is relatively low, that is, the relative importance is relatively low. Figure 3 The relative importance index values ​​of the input and output variables in each industry in the second enterprise screening model are Figure 3It can be seen that among the 8 industries shown in the figure, the trends of the relative importance index values ​​of the variables in the second enterprise screening model are inconsistent. The relative importance of the variable total assets is higher in the power industry, while the relative importance in other industries is lower. Moreover, for the construction industry, the relative importance of all variables is low. From the overall trend of changes in various industries, the trend of changes in the relative importance index values ​​of 6 industries among the 8 industries is basically the same, but the relative importance index values ​​of different industries for the same variable have hierarchical differences. Therefore, from the above analysis, it can be seen that the rationality of the second enterprise screening model is relatively poor when using the original financial variables to calculate the subsequent average efficiency value.

[0125] In this embodiment, the above analysis can be used to obtain the relative importance index values ​​of the input and output variables in the enterprise screening model in each industry, and then the target variable is determined based on the relative importance index values, thereby constructing the third enterprise screening model. Specifically, if there are variables with relative importance index values ​​greater than 0.4 in 4 or more industries among the 8 industries, then the variable can be determined as the target variable. Figure 2 It can be seen that there are input variables P1, P2, P3, P4, P5 and output variables P7, P8, P 10 , P 14 The conditions are met. Through the screening of variables, a total of 9 variables are obtained to construct the third model input and output variable system, and then the third enterprise screening model is constructed based on the third model input and output variable system. The third model input and output variable system is expressed as follows:

[0126] ;

[0127] Among them, the relative importance index values ​​corresponding to each variable in the model input and output variables in the third enterprise screening model are shown in Table 3 below:

[0128] Table 3

[0129]

[0130] Furthermore, in this embodiment, the average of the efficiency values ​​of the first enterprise screening model, the second enterprise screening model, and the third enterprise screening model for evaluating the enterprises to be invested in each industry is calculated. Figure 4 The following table shows the average efficiency of each industry based on three enterprise screening models. For the first enterprise screening model, the formula for the average efficiency of each industry is as follows:

[0131] ;

[0132] represents the mean efficiency of all DMUs in the industry based on the first enterprise screening model in 35 periods. Correspondingly, the mean efficiency of each industry based on the first enterprise screening model, the second enterprise screening model and the third enterprise screening model are .

[0133] pass Figure 5 It can be seen that the mean efficiency of each industry based on the first enterprise screening model is higher than that of the other two models in 8 industries. The mean efficiency of each industry based on the first enterprise screening model and the third enterprise screening model is not much different. Therefore, it can be concluded that the rationality of the mean efficiency of each industry based on the third enterprise screening model is better than that of the first enterprise screening model. For the second enterprise screening model, its mean efficiency of each industry is the lowest. Except for the mining industry, the mean efficiency of each industry in other industries is lower than 0.5. Obviously, it is unreasonable to use the mean efficiency of each industry in this model to screen investment enterprises.

[0134] Furthermore, in this embodiment, by Figure 5 It can be seen that the proportion of industries with an efficiency mean of 1 based on the first enterprise screening model is the highest. The efficiency mean of 6 out of 8 industries is 1, and the proportion of 6 industries with an efficiency mean of 1 is higher than 0.4, indicating that the number of input and output variables in the first enterprise screening model is too large, which makes the efficiency mean of most of the enterprises to be invested 1, which leads to insufficient differentiation of the advantages and disadvantages of the investment enterprises. Therefore, it is not reasonable to screen the investment enterprises based on the first enterprise screening model. In the second enterprise screening model, the proportion of industries with an efficiency mean of 1 is lower than 0.2 in 6 out of 8 industries, and the proportion of industries with an efficiency mean of 1 for electricity is even lower, close to 0. Obviously, it is also unreasonable to screen the investment enterprises based on the second enterprise screening model. For the third enterprise screening model, the proportion of industries with an efficiency mean of 1 is moderate, and there is only one industry with a proportion exceeding 0.5, and the proportion of the remaining 7 industries is lower than 0.4. Therefore, it is more appropriate to screen the investment enterprises based on the third enterprise screening model.

[0135] In this embodiment, the third enterprise screening model is finally used to screen the enterprises to be invested. Specifically, it is determined whether the average efficiency value corresponding to the enterprises to be invested obtained by the third enterprise screening model is greater than 0.9. If the average efficiency value corresponding to the enterprises to be invested is greater than 0.9, the corresponding enterprises to be invested are output as screening results. The following Table 4 shows the relevant information of the enterprises to be invested with an average efficiency value greater than 0.9.

[0136] Table 4

[0137]

[0138] Finally, according to the preset selection rules, the selection results are obtained. In the first enterprise screening model, 192 enterprises to be invested meet the conditions; in the second enterprise screening model, 18 enterprises to be invested meet the conditions; in the third enterprise screening model, 97 enterprises to be invested meet the conditions and these enterprises to be invested are also included in the first enterprise screening model.

[0139] It should be noted that, in this embodiment, there can be a certain degree of flexibility in the second screening of the companies to be invested in the eight industries. In a specific implementation, the target model determined during the first screening can be directly used to screen the companies. This is because a series of rigorous steps have been taken during the first screening process, including obtaining financial data, preprocessing, building models, conducting efficiency evaluation, and model analysis and screening. If the conditions during the first screening and the overall situation of the companies to be invested have not changed significantly, then the target model determined previously may still have good applicability and can re-screen the companies to be invested more reliably and quickly.

[0140] In another specific implementation, the entire process of the first enterprise screening can also be re-executed. Considering that the operating conditions, market environment, industry trends and other factors of the enterprise to be invested may have changed during the two screenings, these changes may cause the previously determined target model to no longer be fully applicable to the current situation, so the complete process of enterprise screening needs to be re-executed.

[0141] Accordingly, see Figure 6 As shown, the embodiment of the present application also provides an enterprise screening device, including:

[0142] The data preprocessing module 11 is used to obtain financial data corresponding to the enterprises to be invested in a number of industries, and preprocess the financial data based on preset data preprocessing rules to obtain corresponding preprocessed data;

[0143] A first model building module 12 is used to determine a first model input-output variable system and a second model input-output variable system by respectively analyzing the financial statement data and the original financial variables in the pre-processed data corresponding to the enterprise to be invested, and to respectively build a first enterprise screening model and a second enterprise screening model based on a preset data envelopment analysis technology and the first model input-output variable system and the second model input-output variable system;

[0144] A second model building module 13 is used to determine the relative importance index values ​​corresponding to the variables in the first model input-output variable system and the second model input-output variable system based on the financial data, and to build a third enterprise screening model based on the preset data envelopment analysis technology and the target variables determined by using the relative importance index values;

[0145] An efficiency evaluation module 14, used to perform efficiency evaluation on the preprocessed data based on the first enterprise screening model, the second enterprise screening model and the third enterprise screening model, respectively, to obtain corresponding first evaluation results, second evaluation results and third evaluation results;

[0146] The enterprise screening module 15 is used to perform model analysis and screening based on the first evaluation result, the second evaluation result and the third evaluation result, and to screen the enterprises to be invested using the target evaluation results corresponding to the determined target enterprise screening model and the preset efficiency threshold to obtain corresponding screening results.

[0147] As can be seen from the above, in this application, the financial data of the enterprise to be invested is first obtained and preprocessed. The preprocessed data is analyzed to determine the input and output variable systems of the first and second models, and the first and second enterprise screening models are constructed. Then the relative importance index values ​​of the variables are determined based on the financial data, and the third enterprise screening model is constructed. The data are evaluated for efficiency using three models, and model analysis and screening are performed based on the results to determine the target enterprise screening model. Finally, the target evaluation results and the preset efficiency threshold are used to screen the enterprises to be invested, and the screening results are obtained. In this way, this application can effectively improve the accuracy and reliability of enterprise screening for users to make investment references.

[0148] In a specific implementation, the data preprocessing module 11 may specifically include:

[0149] A first data acquisition unit, configured to remove data that meets a first preset data missing condition in the financial data, and to interpolate missing values ​​for data that meets a second preset data missing condition in the financial data, so as to complete a first data processing operation and obtain corresponding first data;

[0150] A first judgment result obtaining unit, configured to judge whether there is target data in the first data that meets a preset extreme value condition based on the average value and the standard deviation respectively corresponding to each financial variable in the first data, so as to obtain a corresponding first judgment result;

[0151] The second data acquisition unit is used to remove the target data if the first judgment result indicates that the target data exists, so as to complete the second data processing operation and obtain pre-processed data.

[0152] In a specific implementation, the first model building module 12 may specifically include:

[0153] A first system construction unit, configured to construct a first model input and output variable system by analyzing the financial performance indicators in the financial statement data in the preprocessed data;

[0154] The second system building unit is used to build a second model input and output variable system by analyzing the original financial variables in the preprocessed data.

[0155] In a specific implementation, the enterprise screening module 15 may specifically include:

[0156] an analysis result acquisition unit, for determining, based on the first evaluation result, the second evaluation result and the third evaluation result, the average of the efficiency values ​​of the first enterprise screening model, the second enterprise screening model and the third enterprise screening model for evaluating the enterprises to be invested in each industry, and performing model analysis according to the obtained average efficiency values ​​of each industry to obtain corresponding model analysis results;

[0157] A model screening unit is used to screen out a target enterprise screening model from the first enterprise screening model, the second enterprise screening model and the third enterprise screening model based on the model analysis result.

[0158] In a specific implementation, the enterprise screening module 15 may specifically include:

[0159] An evaluation result acquisition unit, used to acquire a target evaluation result corresponding to the determined target enterprise screening model;

[0160] A second judgment result obtaining unit, used to determine an average efficiency value corresponding to the enterprise to be invested based on the target evaluation result, and to determine whether the average efficiency value is greater than a preset efficiency threshold, so as to obtain a corresponding second judgment result;

[0161] A screening result acquisition unit is used to screen the enterprises to be invested in based on the second judgment result to obtain corresponding screening results.

[0162] In a specific implementation, the second model building module 13 may specifically include:

[0163] A weight value determination unit, used to determine, based on the financial data, the relative weight values ​​of the variables in the first model input-output variable system and the second model input-output variable system in each time period and in each industry;

[0164] The index value acquisition unit is used to calculate the average value of the relative weight value of each variable based on the corresponding number of time periods to obtain the relative importance index value of each variable in each industry.

[0165] In a specific implementation, the second model building module 13 may specifically include:

[0166] A third judgment result obtaining unit, used to respectively judge whether the relative importance index value of each variable in each industry is not less than a preset importance threshold value, so as to obtain a corresponding third judgment result;

[0167] a fourth judgment result obtaining unit, configured to count the target number of the third judgment results corresponding to each of the variables and having a representation not less than the preset importance threshold, and to determine whether each of the target numbers is not less than the target threshold, so as to obtain a corresponding fourth judgment result;

[0168] A model building unit is used to determine the corresponding variable as a target variable if the fourth judgment result indicates that it is not less than the target threshold, and to build a third enterprise screening model based on the preset data envelopment analysis technology and the target variable.

[0169] Furthermore, the present application also discloses an electronic device. Figure 7 This is a structural diagram of an electronic device 20 according to an exemplary embodiment, and the content in the figure cannot be regarded as any limitation on the scope of use of this application. The electronic device 20 may specifically include: at least one processor 21, at least one memory 22, a power supply 23, a communication interface 24, an input / output interface 25, and a communication bus 26. The memory 22 is used to store a computer program, which is loaded and executed by the processor 21 to implement the relevant steps in the enterprise screening method disclosed in any of the aforementioned embodiments. In addition, the electronic device 20 in this embodiment may specifically be an electronic computer.

[0170] In this embodiment, the power supply 23 is used to provide working voltage for each hardware device on the electronic device 20; the communication interface 24 can create a data transmission channel between the electronic device 20 and the external device, and the communication protocol it follows is any communication protocol that can be applied to the technical solution of the present application, and is not specifically limited here; the input and output interface 25 is used to obtain external input data or output data to the outside world, and its specific interface type can be selected according to specific application needs and is not specifically limited here.

[0171] In addition, the memory 22, as a carrier for storing resources, can be a read-only memory, a random access memory, a disk or an optical disk, etc. The resources stored thereon can include an operating system 221, a computer program 222, etc., and the storage method can be temporary storage or permanent storage.

[0172] The operating system 221 is used to manage and control the hardware devices and computer programs 222 on the electronic device 20, and can be Windows Server, Netware, Unix, Linux, etc. In addition to the computer program that can be used to complete the enterprise screening method performed by the electronic device 20 disclosed in any of the aforementioned embodiments, the computer program 222 can further include computer programs that can be used to complete other specific tasks.

[0173] Furthermore, the present application also discloses a computer-readable storage medium for storing a computer program; wherein the computer program, when executed by a processor, implements the enterprise screening method disclosed above. The specific steps of the method can refer to the corresponding contents disclosed in the above embodiments, and will not be repeated here.

[0174] In this specification, each embodiment is described in a progressive manner, and each embodiment focuses on the differences from other embodiments. The same or similar parts between the embodiments can be referred to each other. For the device disclosed in the embodiment, since it corresponds to the method disclosed in the embodiment, the description is relatively simple, and the relevant parts can be referred to the method part.

[0175] Professionals may further appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, computer software, or a combination of the two. In order to clearly illustrate the interchangeability of hardware and software, the composition and steps of each example have been generally described in the above description according to function. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professionals and technicians may use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of this application.

[0176] The steps of the method or algorithm described in conjunction with the embodiments disclosed herein may be implemented directly using hardware, a software module executed by a processor, or a combination of the two. The software module may be placed in a random access memory (RAM), a memory, a read-only memory (ROM), an electrically programmable ROM, an electrically erasable programmable ROM, a register, a hard disk, a removable disk, a CD-ROM, or any other form of storage medium known in the art.

[0177] Finally, it should be noted that, in this article, relational terms such as first and second, etc. are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "include", "comprise" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device. In the absence of further restrictions, the elements defined by the sentence "comprise a ..." do not exclude the presence of other identical elements in the process, method, article or device including the elements.

[0178] The technical solution provided by the present application is introduced in detail above. Specific examples are used in this article to illustrate the principles and implementation methods of the present application. The description of the above embodiments is only used to help understand the method of the present application and its core idea. At the same time, for general technicians in this field, according to the idea of ​​the present application, there will be changes in the specific implementation methods and application scope. In summary, the content of this specification should not be understood as a limitation on the present application.

Claims

1. A method for screening enterprises, characterized in that: include: Acquire financial data corresponding to the enterprises to be invested in several industries, and pre-process the financial data based on preset data pre-processing rules to obtain corresponding pre-processed data; By respectively analyzing the financial statement data and the original financial variables in the pre-processed data corresponding to the enterprise to be invested, a first model input-output variable system and a second model input-output variable system are determined, and a first enterprise screening model and a second enterprise screening model are respectively constructed based on a preset data envelopment analysis technology and the first model input-output variable system and the second model input-output variable system; Determine the relative importance index values ​​corresponding to the variables in the first model input-output variable system and the second model input-output variable system based on the financial data, and construct a third enterprise screening model based on the preset data envelopment analysis technology and the target variables determined by using the relative importance index values; Respectively performing efficiency evaluation on the preprocessed data based on the first enterprise screening model, the second enterprise screening model, and the third enterprise screening model to obtain corresponding first evaluation results, second evaluation results, and third evaluation results; Model analysis and screening are performed based on the first evaluation result, the second evaluation result and the third evaluation result, and the target evaluation results corresponding to the determined target enterprise screening model and the preset efficiency threshold are used to screen the enterprises to be invested in order to obtain corresponding screening results.

2. The enterprise screening method according to claim 1, characterized in that: The preprocessing of the financial data based on the preset data preprocessing rules to obtain corresponding preprocessed data includes: Eliminate data that meets a first preset data missing condition in the financial data, and perform missing value interpolation on data that meets a second preset data missing condition in the financial data, so as to complete a first data processing operation and obtain corresponding first data; Based on the average values ​​and standard deviations respectively corresponding to the financial variables in the first data, it is judged whether there is target data in the first data that meets the preset extreme value condition, so as to obtain a corresponding first judgment result; If the first judgment result indicates that the target data exists, the target data is eliminated to complete the second data processing operation and obtain pre-processed data.

3. The enterprise screening method according to claim 1, characterized in that: The method of respectively analyzing the financial statement data and the original financial variables in the pre-processed data corresponding to the enterprise to be invested in, so as to determine the first model input and output variable system and the second model input and output variable system, comprises: By analyzing the financial performance indicators in the financial statement data in the pre-processed data, a first model input and output variable system is constructed; The input and output variable system of the second model is constructed by analyzing the original financial variables in the preprocessed data.

4. The enterprise screening method according to claim 1, characterized in that: The performing of model analysis and screening based on the first evaluation result, the second evaluation result, and the third evaluation result includes: Determine the mean of the efficiency values ​​of the first enterprise screening model, the second enterprise screening model and the third enterprise screening model for evaluating the enterprises to be invested in each industry based on the first evaluation result, the second evaluation result and the third evaluation result, and perform model analysis based on the obtained mean efficiency values ​​of each industry to obtain corresponding model analysis results; A target enterprise screening model is screened out from the first enterprise screening model, the second enterprise screening model and the third enterprise screening model based on the model analysis result.

5. The enterprise screening method according to claim 1, characterized in that: The method of screening the enterprises to be invested by using the target evaluation results corresponding to the determined target enterprise screening model and the preset efficiency threshold to obtain corresponding screening results includes: Obtaining target evaluation results corresponding to the determined target enterprise screening model; Determine an average efficiency value corresponding to the enterprise to be invested based on the target evaluation result, and judge whether the average efficiency value is greater than a preset efficiency threshold to obtain a corresponding second judgment result; Based on the second judgment result, the enterprises to be invested are screened to obtain corresponding screening results.

6. The enterprise screening method according to any one of claims 1 to 5, characterized in that: The determining, based on the financial data, the relative importance index values ​​corresponding to the variables in the first model input-output variable system and the second model input-output variable system, respectively, includes: Determine, based on the financial data, the relative weight values ​​of the variables in the first model input-output variable system and the second model input-output variable system in each time period and in each industry; The average value of the relative weight value of each variable is calculated based on the corresponding number of time periods to obtain the relative importance index value of each variable in each industry.

7. The enterprise screening method according to claim 6, characterized in that: The method of constructing a third enterprise screening model based on the preset data envelopment analysis technology and the target variables determined by using the relative importance index values ​​includes: respectively judging whether the relative importance index value of each of the variables in each of the industries is not less than a preset importance threshold, so as to obtain a corresponding third judgment result; Counting the target number of the third judgment result corresponding to each of the variables and representing a value not less than the preset importance threshold, and judging whether each of the target numbers is not less than the target threshold, so as to obtain a corresponding fourth judgment result; If the fourth judgment result indicates that it is not less than the target threshold, the corresponding variable is determined to be the target variable, and a third enterprise screening model is constructed based on the preset data envelopment analysis technology and the target variable.

8. An enterprise screening device, characterized in that: include: A data preprocessing module, used to obtain financial data corresponding to the enterprises to be invested in several industries, and preprocess the financial data based on preset data preprocessing rules to obtain corresponding preprocessed data; A first model building module is used to determine a first model input-output variable system and a second model input-output variable system by respectively analyzing the financial statement data and the original financial variables in the pre-processed data corresponding to the enterprise to be invested, and to respectively build a first enterprise screening model and a second enterprise screening model based on a preset data envelopment analysis technology and the first model input-output variable system and the second model input-output variable system; A second model building module is used to determine the relative importance index values ​​corresponding to the variables in the first model input-output variable system and the second model input-output variable system based on the financial data, and to build a third enterprise screening model based on the preset data envelopment analysis technology and the target variables determined by using the relative importance index values; an efficiency evaluation module, used to perform efficiency evaluation on the preprocessed data based on the first enterprise screening model, the second enterprise screening model, and the third enterprise screening model, respectively, to obtain corresponding first evaluation results, second evaluation results, and third evaluation results; An enterprise screening module is used to perform model analysis and screening based on the first evaluation result, the second evaluation result and the third evaluation result, and to screen the enterprises to be invested using the target evaluation results corresponding to the determined target enterprise screening model and a preset efficiency threshold to obtain corresponding screening results.

9. An electronic device, characterized in that: include: Memory, used to store computer programs; A processor, configured to execute the computer program to implement the enterprise screening method according to any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that: Used to store a computer program; wherein, when the computer program is executed by a processor, the enterprise screening method according to any one of claims 1 to 7 is implemented.

Images