Enterprise investment efficiency measuring and calculating system based on LASSO-game theory combination empowerment-super-efficiency SBM

Abstract

The invention provides an enterprise investment efficiency measuring and calculating system based on LASSO-game theory combined empowerment-super efficiency SBM, and the system comprises the steps: firstly, carrying out the screening of an existing input-output index system based on an LASSO method, and further providing subjective intervention in an evaluation system construction process; and then, combining output index weights obtained by the entropy weight method and the AHP method by adopting a game theory combined weighting method, and obtaining an output comprehensive index by combining a linear weighting method, thereby realizing scientific and reasonable weighting and dimension reduction of the output index. And finally, measuring and calculating the investment efficiency of the enterprise by adopting a super-efficiency SBM model considering unexpected output. According to the system provided by the invention, redundant information in an evaluation system is effectively simplified, estimation deviation caused by index dimension and importance difference is avoided, and unexpected output is introduced into a calculation process of an enterprise investment efficiency level, so that a result has more practical significance.

Description

Technical Field

[0001] The present invention relates to the technical field of enterprise investment efficiency calculation, and in particular to an enterprise investment efficiency calculation system based on LASSO-game theory combined weighting-super efficiency SBM. Background Art

[0002] Improving the scientific nature of enterprise investment efficiency measurement is an effective path to objectively assessing corporate sustainable development. Current methods for estimating enterprise investment efficiency, on the one hand, generally overlook the potential undesirable outcomes of investment activities, such as environmental pollution and carbon emissions, failing to truly reflect their impact on society and the environment. On the other hand, they fail to fully account for the differences in the scale and importance of output indicators. Furthermore, existing methods for weighting and dimensionality reduction indicators often employ a single weighting approach, potentially leading to an imbalance in weight distribution and compromising the scientific and rationale of the results. Currently, there is a lack of solutions that simultaneously address these issues. Summary of the Invention

[0003] The purpose of this invention is to propose an enterprise investment efficiency calculation system based on LASSO-game theory combined empowerment-super efficiency SBM, in order to support the long-term efficient and green development of enterprises under the goal of sustainable high-quality development.

[0004] The present invention is based on a computer system and screens the existing indicator system based on the idea of regularization, thereby eliminating redundant variables or variables that contribute little to the target variable and improving interpretability. Then, the game theory combined weighting method is combined with the advantages of subjective and objective weighting methods to obtain scientific and reasonable indicator weights, and the output indicator is reduced in dimension to avoid estimation bias caused by differences in dimension and importance. Finally, the super-efficiency SBM model including non-expected output is used to measure the investment efficiency of the enterprise. The system provided by the present invention not only effectively simplifies the redundant information of the indicator system, but also eliminates the influence of subjective will and data deviation. The investment efficiency level of the enterprise thus calculated is more realistic.

[0005] To achieve the above objectives, the technical solution of the present invention is: a system for calculating enterprise investment efficiency based on LASSO-game theory combined weighting-super efficiency SBM, comprising:

[0006] Evaluation index system construction module, used to construct a multi-layer enterprise investment efficiency evaluation index system, including input indicators, expected output indicators to be standardized and reduced in dimension, and non-expected output indicators;

[0007] The data screening module is used to construct an equation between enterprise investment efficiency evaluation indicators and enterprise performance based on the LASSO algorithm, eliminate redundant or low-correlation indicators, retain core indicators, optimize the input and output indicator sets, and form a streamlined enterprise investment efficiency measurement indicator system;

[0008] The standardization processing module is used to standardize the indicators in the enterprise investment efficiency measurement indicator system after screening based on the LASSO algorithm;

[0009] The weight calculation module is used to combine the entropy weight method and the analytic hierarchy process (AHP) to obtain the objective weight and subjective weight of each output indicator respectively. Based on game theory, the two types of weights are combined to calculate the final weight of each output indicator;

[0010] The output index calculation module is used to calculate the comprehensive output index of each dimension based on the final weight of the output indicator using a linear weighting method;

[0011] The investment efficiency calculation module is used to calculate the investment efficiency of enterprises based on the super-efficiency SBM model, sort and output them in combination with the calculation structure.

[0012] Preferably, the expected output indicators include economic benefits and social benefits, and the undesired output indicators include environmental pollution.

[0013] Preferably, the environmental pollution indicators include carbon dioxide emissions, sulfur dioxide emissions, and smoke emissions.

[0014] Preferably, in the data screening module, the LASSO regression method is used to examine the relationship between the investment efficiency evaluation index and enterprise performance of all enterprise samples; specifically, as follows:

[0015] Assume that in the evaluation index system construction module, there are R indicators in the enterprise investment efficiency evaluation index system constructed; there are M groups of enterprise samples, and each group of enterprise samples includes R evaluation indicators and corresponding enterprise performance of enterprises within a preset statistical period;

[0016] Building a regression model where y i is the enterprise performance corresponding to the i-th group of enterprise samples, x ir is the rth evaluation index in the i-th group of enterprise samples, β r is the regression coefficient corresponding to the rth indicator; β0 is the constant term; ε i is a random error;

[0017] Optimization objective based on LASSO regression Select the penalty coefficient λ that minimizes the cross-validation error to determine the optimal parameter;

[0018] The optimal penalty coefficient is used to fit the LASSO regression, the optimal regression coefficient is solved, and the indicators with a regression coefficient of 0 are eliminated to obtain a new indicator system. At this time, the evaluation system contains P input indicators and Q output indicators, among which the expected output indicators include q1 economic benefit indicators, q2 social benefit indicators, and q3 non-expected output indicators.

[0019] Preferably, data standardization is performed on the expected output index and the non-expected output index in the selected enterprise investment efficiency measurement index system. The original data is X=(x ij ) M×Q (i=1,2,…M;j=1,2,…Q); where x ij represents the j-th output indicator in the i-th group of enterprise samples; the standardized processing formula is:

[0020] Preferably, in the weight calculation module, the entropy weight method is used to determine the weight of the output indicator in the investment efficiency calculation index system of the screened enterprise. The calculation process is as follows:

[0021] First, calculate the information entropy of each output dimension: in When p ij = 0, p ij lnp ij =0;

[0022] Secondly, calculate the entropy weight of the j-th indicator: in,

[0023] Preferably, in the weight calculation module, the calculation process of determining the output index weights in the enterprise investment efficiency measurement index system after screening using the hierarchical analysis method is as follows:

[0024] First, construct a discriminant matrix. For the factors in the same layer of the enterprise investment efficiency measurement index system that belong to each factor in the upper layer after screening, use the paired comparison method and the 1-9 comparison scale to construct a paired comparison matrix until the bottom layer; obtain the discriminant matrix A = (a kl ) n×n ,k,l=1,2,...n, where a kl Indicates the importance of factor k relative to factor l. Both k and l represent the factors of the same layer that belong to each factor of the previous layer. n represents the number of factors. a kk =1,a kl =a lk ;

[0025] Second, the characteristic equation of the discriminant matrix is used to determine the eigenvector, and the eigenvector is normalized to obtain the weight vector W k , determine the maximum eigenvalue

[0026] Third, calculate the consistency index CI = 0 indicates perfect agreement, and the larger the CI, the less agreement;

[0027] Fourth, calculate the consistency ratio RI is the random consistency index. When CR is less than 0.1, the matrix is consistent and the discriminant matrix is reasonable.

[0028] Finally, the relative weight values of each layer are calculated from bottom to top and multiplied to obtain the indicator weight under the AHP method.

[0029] Preferably, in the weight calculation module, the two types of weights are combined based on game theory to calculate the final weight of each output indicator. The specific process is as follows:

[0030] Treat the selected output indicators as the player set N in game theory;

[0031] Calculate the Shapley value of indicator j Where |N| represents the number of elements in the set of indicators, S is any subset that does not contain j, |S| is the number of participants in subset S, v(S) is the value of the cooperative combination S, that is, the sum of the weights of all indicators in subset S under the entropy weight method and the AHP method, v(S∪{j})-v(S) is the marginal contribution of indicator j in subset S, The weight coefficient of all permutations of indicator j when adding it to subset S to ensure fair weighting;

[0032] Multiply the Shapley value of each indicator by the sum of the weights of its two weighting methods to obtain the combined weight ω of the indicator j C =V(j)×(ω j E +ω j A ), where ω j E and ω j A The two sets of original weight vectors provided by the entropy weight method and the AHP method are respectively; then, all indicator weights are normalized to ensure that the sum of the weights is 1, and the final combined weight of the jth indicator can be obtained

[0033] Preferably, the final weight obtained by the game theory combined weighting method is used to perform weighted summation on the output indicators of each dimension to obtain the expected output comprehensive index g and the unexpected output comprehensive index h.

[0034] Preferably, in the investment efficiency calculation module, a super-efficiency SBM model including non-expected output is used to calculate the investment efficiency index of enterprises and then sort them; the super-efficiency SBM model with a BCC structure is:

[0035]

[0036] Among them, δ represents the efficiency value of the decision-making unit, subscript 0 is the decision-making unit being evaluated, subscript i represents the i-th enterprise sample, f z0 , g0 and h0 represent the input value, expected output value and unexpected output value of the evaluated decision-making unit 0 on the z-th input indicator respectively; P represents the number of input indicators, f zi 、g i and h i represents the value of the input index, expected output index and unexpected output index of enterprise sample i at the zth time, are the redundancy, expected output deficiency and unexpected output superscalar matrices of the z-th input of the decision-making unit, γ i and γ zi represents the weight variable;

[0037] If the CCR structure super-efficiency SBM model is selected, remove and This condition;

[0038] For the final efficiency value, if δ>1, it indicates that the enterprise is super-efficient; if δ=1, it indicates that the enterprise is just at the efficient frontier; if δ<1, it indicates that the enterprise needs to improve efficiency.

[0039] Compared with the prior art, the present invention has the following beneficial effects:

[0040] The calculation system provided by the present invention and its preferred embodiment not only selects indicators important to enterprise investment efficiency during the enterprise investment efficiency measurement process, thus avoiding interference from redundant variables, but also integrates the advantages of subjective and objective weights when weighting and reducing the output index, making the weight distribution more scientific and reasonable, and effectively avoiding the efficiency estimation bias caused by differences in the dimensions and importance of different indicators. In addition, this system fully considers the impact of undesirable outputs in the enterprise investment efficiency measurement, making the calculated enterprise investment efficiency level more consistent with the actual economic environment and more practical. It not only helps enterprises optimize their investment decisions, but also provides a valuable reference for achieving sustainable economic development. BRIEF DESCRIPTION OF THE DRAWINGS

[0041] Figure 1 This is a system workflow diagram of an embodiment of the present invention. DETAILED DESCRIPTION

[0042] The technical solution of the present invention will be described in detail below with reference to the accompanying drawings.

[0043] In order to make the purpose, technical solutions and advantages of this application more clearly understood, the present application is described and illustrated below in conjunction with the accompanying drawings and examples. It should be understood that the specific embodiments described herein are merely used to explain this application and are not intended to limit this application. Based on the embodiments provided in this application, all other embodiments obtained by those of ordinary skill in the art without making any creative efforts are within the scope of protection of this application.

[0044] Obviously, the drawings described below are merely examples or embodiments of the present application. Those skilled in the art can, without inventive effort, apply the present application to other similar scenarios based on these drawings. Furthermore, it is also understood that, although the effort involved in such a development process may be complex and lengthy, for those skilled in the art related to the content disclosed in this application, changes in design, manufacturing, or production based on the technical content disclosed in this application are merely conventional technical means and should not be construed as an insufficiency of the content disclosed in this application.

[0045] References to "embodiments" in this application mean that a particular feature, structure, or characteristic described in connection with the embodiment may be included in at least one embodiment of the application. The appearance of this phrase in various places in the specification does not necessarily refer to the same embodiment, nor does it refer to independent or alternative embodiments that are mutually exclusive of other embodiments. It is understood, both explicitly and implicitly, by those skilled in the art that the embodiments described in this application may be combined with other embodiments unless there is a conflict.

[0046] Unless otherwise defined, the technical or scientific terms used in this application should have the ordinary meaning understood by a person of ordinary skill in the technical field to which this application belongs. The words "one", "a", "the" and the like used in this application do not indicate a limit on quantity and may indicate the singular or plural. The terms "include", "comprise", "have" and any variations thereof used in this application are intended to cover non-exclusive inclusions; for example, a process, method, system, product or device that includes a series of steps or modules (units) is not limited to the listed steps or units, but may also include steps or units that are not listed, or may also include other steps or units that are inherent to these processes, methods, products or devices. The words "connect", "connected", "coupled" and the like used in this application are not limited to physical or mechanical connections, but may include electrical connections, whether direct or indirect. The word "multiple" used in this application refers to two or more. "And / or" describes the association relationship of associated objects, indicating that three relationships can exist. For example, "A and / or B" can mean: A exists alone, A and B exist at the same time, and B exists alone. The character " / " generally indicates that the objects before and after are in an "or" relationship. The terms "first", "second", "third", etc. involved in this application are only used to distinguish similar objects and do not represent a specific order for the objects.

[0047] In an embodiment of the present invention, a system for calculating enterprise investment efficiency based on LASSO-game theory combined weighting-super efficiency SBM is provided, comprising:

[0048] Evaluation index system construction module, used to construct a multi-layer enterprise investment efficiency evaluation index system, including input indicators, expected output indicators to be standardized and reduced in dimension, and non-expected output indicators;

[0049] The data screening module is used to construct an equation between enterprise investment efficiency evaluation indicators and enterprise performance based on the LASSO algorithm, eliminate redundant or low-correlation indicators, retain core indicators, optimize the input and output indicator sets, and form a streamlined enterprise investment efficiency measurement indicator system;

[0050] The standardization processing module is used to standardize the indicators in the enterprise investment efficiency measurement indicator system after screening based on the LASSO algorithm;

[0051] The weight calculation module is used to combine the entropy weight method and the analytic hierarchy process (AHP) to obtain the objective weight and subjective weight of each output indicator respectively. Based on game theory, the two types of weights are combined to calculate the final weight of each output indicator;

[0052] The output index calculation module is used to calculate the comprehensive output index of each dimension based on the final weight of the output indicator using a linear weighting method;

[0053] The investment efficiency calculation module is used to calculate the investment efficiency of enterprises based on the super-efficiency SBM model, sort and output them in combination with the calculation structure.

[0054] like Figure 1 As shown, its workflow specifically includes the following steps:

[0055] Step S1: Determine the enterprise sample, build an enterprise investment efficiency measurement indicator system, and collect data.

[0056] Step S2: Based on the obtained data, the LASSO regression method is used to examine the relationship between the investment efficiency evaluation indicators and corporate performance of all enterprise samples, and to screen the indicators.

[0057] Step S3: Perform data standardization on the expected output and unexpected output indicators in the screened evaluation system.

[0058] Step S4: Use the entropy weight method and AHP method to determine the subjective and objective weights of the output indicators in the enterprise investment efficiency evaluation system.

[0059] Step S5: Determine the final weight of the output indicator based on the Shapley value concept of game theory.

[0060] Step S6: Use the final weight of the output indicator to perform weighted summation on the output indicators of each dimension to obtain a comprehensive index of expected output and unexpected output.

[0061] Step S7: Calculate the enterprise investment efficiency index using the super-efficiency SBM model that includes non-expected output, and then sort the index.

[0062] In this embodiment, the desired output indicators include economic benefits and social benefits, and the undesired output indicators include environmental pollution.

[0063] In this embodiment, the environmental pollution indicators include carbon dioxide emissions, sulfur dioxide emissions, and smoke emissions.

[0064] In this embodiment, the data screening module uses the LASSO regression method to examine the relationship between the investment efficiency evaluation index and enterprise performance of all enterprise samples; the details are as follows:

[0065] Assume that in the evaluation index system construction module, there are R indicators in the enterprise investment efficiency evaluation index system constructed; there are M groups of enterprise samples, and each group of enterprise samples includes R evaluation indicators and corresponding enterprise performance of enterprises within a preset statistical period;

[0066] Building a regression model where y i is the enterprise performance corresponding to the i-th group of enterprise samples, xir is the rth evaluation index in the i-th group of enterprise samples, β r is the regression coefficient corresponding to the rth indicator; β0 is the constant term; ε i is a random error;

[0067] Optimization objective based on LASSO regression Select the penalty coefficient λ that minimizes the cross-validation error to determine the optimal parameter;

[0068] The optimal penalty coefficient is used to fit the LASSO regression, the optimal regression coefficient is solved, and the indicators with a regression coefficient of 0 are eliminated to obtain a new indicator system. At this time, the evaluation system contains P input indicators and Q output indicators, among which the expected output indicators include q1 economic benefit indicators, q2 social benefit indicators, and q3 non-expected output indicators.

[0069] In this embodiment, data standardization is performed on the expected output index and the non-expected output index in the selected enterprise investment efficiency calculation index system. The original data is X=(x ij ) M×Q (i=1,2,...M;j=1,2,...Q); where x ij represents the j-th output indicator in the i-th group of enterprise samples; the standardized processing formula is:

[0070] In this embodiment, in the weight calculation module, the entropy weight method is used to determine the output index weights in the enterprise investment efficiency measurement index system after screening. The calculation process is as follows:

[0071] First, calculate the information entropy of each output dimension: in When p ij = 0, p ij lnp ij =0;

[0072] Secondly, calculate the entropy weight of the j-th indicator: in,

[0073] In this embodiment, in the weight calculation module, the calculation process of determining the output indicator weights in the enterprise investment efficiency measurement indicator system after screening using the hierarchical analysis method is as follows:

[0074] First, construct a discriminant matrix. For the factors in the same layer of the enterprise investment efficiency measurement index system that belong to each factor in the upper layer after screening, use the paired comparison method and the 1-9 comparison scale to construct a paired comparison matrix until the bottom layer; obtain the discriminant matrix A = (a kl ) n×n ,k,l=1,2,…n, where akl Indicates the importance of factor k relative to factor l. Both k and l represent the factors of the same layer that belong to each factor of the previous layer. n represents the number of factors. a kk =1,a kl =a lk ;

[0075] Second, the characteristic equation of the discriminant matrix is used to determine the eigenvector, and the eigenvector is normalized to obtain the weight vector W k , determine the maximum eigenvalue

[0076] Third, calculate the consistency index CI = 0 indicates perfect agreement, and the larger the CI, the less agreement;

[0077] Fourth, calculate the consistency ratio RI is the random consistency index. When CR is less than 0.1, the matrix is consistent and the discriminant matrix is reasonable.

[0078] Finally, the relative weight values of each layer are calculated from bottom to top and multiplied to obtain the indicator weight under the AHP method.

[0079] In this embodiment, the weight calculation module combines two types of weights based on game theory to calculate the final weights of various output indicators. The specific process is as follows:

[0080] The Shapley value is a method of profit distribution in cooperative games. The core idea is to calculate the average marginal contribution of each individual in all possible cooperation sequences to ensure that the final distribution complies with the principle of fairness.

[0081] Treat the selected output indicators as the player set N in game theory;

[0082] Calculate the Shapley value of indicator j Where |N| represents the number of elements in the set of indicators, S is any subset that does not contain j, |S| is the number of participants in subset S, v(S) is the value of the cooperative combination S, that is, the sum of the weights of all indicators in subset S under the entropy weight method and the AHP method, v(S∪{j})-v(S) is the marginal contribution of indicator j in subset S, The weight coefficient of all permutations of indicator j when adding it to subset S to ensure fair weighting;

[0083] Multiply the Shapley value of each indicator by the sum of the weights of its two weighting methods to obtain the combined weight ω of the indicator j C =V(j)×(ω j E +ω j A ), where ωj E and ω j A These are the two sets of original weight vectors provided by the entropy weight method and the AHP method respectively;

[0084] Then, all indicator weights are normalized to ensure that the sum of the weights is 1, and the final combined weight of the jth indicator can be obtained.

[0085] In this embodiment, the final weight obtained by the game theory combined weighting method is used to perform weighted summation on the output indicators of each dimension to obtain the expected output comprehensive index g and the unexpected output comprehensive index h.

[0086] In this embodiment, the investment efficiency calculation module uses a super-efficiency SBM model that includes undesirable output to calculate the investment efficiency index of enterprises and then rank them; the super-efficiency SBM model with a BCC structure is:

[0087]

[0088] Among them, δ represents the efficiency value of the decision-making unit, subscript 0 is the decision-making unit being evaluated, subscript i represents the i-th enterprise sample, f z0 , g0 and h0 represent the input value, expected output value and unexpected output value of the evaluated decision-making unit 0 on the z-th input indicator respectively; P represents the number of input indicators, f zi 、g i and h i represents the value of the input index, expected output index and unexpected output index of enterprise sample i at the zth time, are the redundancy, expected output deficiency and unexpected output superscalar matrices of the z-th input of the decision-making unit, γ i and γ zi represents the weight variable;

[0089] If the CCR structure super-efficiency SBM model is selected, remove and This condition;

[0090] For the final efficiency value, if δ>1, it indicates that the enterprise is super-efficient; if δ=1, it indicates that the enterprise is just at the efficient frontier; if δ<1, it indicates that the enterprise needs to improve efficiency.

[0091] The following uses a specific application example to further illustrate the working process and specific processing steps of the present system. In this example, Step S1 specifically involves constructing an enterprise investment efficiency evaluation system for a sample area of 10 enterprises, totaling 60 samples, and providing detailed data descriptions. Considering the high privacy nature of enterprise data, random numbers are generated based on the properties of each indicator for simulation purposes. The evaluation system is shown in Table 1, and the relevant data is provided in Tables 2-1, 2-2, and 2-3.

[0092] Table 1 Enterprise investment efficiency index system

[0093]

[0094]

[0095] Table 2-1 Original data

[0096]

[0097]

[0098]

[0099] Table 2-2 Original data

[0100]

[0101]

[0102]

[0103] Table 2-3 Original data

[0104]

[0105]

[0106]

[0107] In this embodiment, the specific content of step S2 is: based on the LASSO regression equation, the relationship between each indicator and enterprise performance is examined. The specific results are shown in Table 3:

[0108] Table 3 Index coefficient results based on LASSO regression

[0109] Indicator name coefficient Indicator name coefficient Indicator name coefficient Indicator name coefficient X1 0 X6 5.84 X11 876.63 X16 0 X2 0.37 X7 11.07 X12 1660.09 X17 0.79 X3 0 X8 0.49 X13 36175.40 X18 0 X4 0 X9 2.59 X14 11109.33 X19 0 X5 0 X10 0 X15 0

[0110] Furthermore, in this embodiment, the specific content of step S2 is: based on the regression results of Table 3, the indicators with a coefficient of 0 are eliminated, and a screened indicator system is formed, as shown in Table 4:

[0111] Table 4 Simplified enterprise investment efficiency calculation index system

[0112]

[0113]

[0114] In this embodiment, the specific content of step S3 is: performing data standardization on the simplified expected output and undesired output. The specific results are shown in Table 5:

[0115] Table 5 Standardization

[0116]

[0117]

[0118]

[0119] In this embodiment, the specific content of step S4 is: using the entropy weight method and the AHP method to determine the subjective and objective weights of the output indicators in the enterprise investment efficiency evaluation system. The specific results are shown in Table 6:

[0120] Table 6 Indicator weights

[0121] index Entropy weight method weight (%) AHP method weight (%) X11 15.26 18.65 X12 25.50 15.87 X13 8.68 11.59 X14 23.65 35.55 X17 26.91 18.34

[0122] In this embodiment, the specific content of step S5 is: based on the game theory Shapley value concept, the output index weight is determined as the final weight basis for calculating the output index of each dimension after screening. The specific results are shown in Table 7:

[0123] Table 7 Game theory method to determine the indicator weight coefficient

[0124]

[0125]

[0126] In this embodiment, the specific content of step S6 is: based on the final weight coefficient, the output index of each dimension is weighted and summed to obtain the comprehensive index of expected output and undesired output. The specific results are shown in Table 8:

[0127] Table 8 Expected output and unexpected output index

[0128]

[0129]

[0130]

[0131] In this embodiment, the specific content of step S7 is: using the super-efficiency SBM model to calculate the enterprise investment efficiency index, and then sorting the results. The specific results are shown in Table 9:

[0132] Table 9 Enterprise investment efficiency level and ranking

[0133]

[0134]

[0135]

[0136] The above results eliminate subjective interference and information redundancy from the corporate investment efficiency index, making the results more realistic. The results show that Company 5 has the highest average efficiency value and excellent overall investment efficiency performance, but its ranking fluctuates significantly. Company 8 follows closely behind, achieving the highest efficiency peak among all companies and even achieving the top ranking in some years. Company 1 has a more balanced overall efficiency level and relatively stable investment efficiency.

[0137] The above are preferred embodiments of the present invention. Any changes made according to the technical solution of the present invention, as long as the resulting functions and effects do not exceed the scope of the technical solution of the present invention, shall fall within the scope of protection of the present invention.

Claims

1. A system for calculating enterprise investment efficiency based on LASSO-game theory combined weighting-super-efficiency SBM, characterized by: include: Evaluation index system construction module, used to construct a multi-layer enterprise investment efficiency evaluation index system, including input indicators, expected output indicators to be standardized and reduced in dimension, and non-expected output indicators; The data screening module is used to construct an equation between enterprise investment efficiency evaluation indicators and enterprise performance based on the LASSO algorithm, eliminate redundant or low-correlation indicators, retain core indicators, optimize the input and output indicator sets, and form a streamlined enterprise investment efficiency measurement indicator system; The standardization processing module is used to standardize the indicators in the enterprise investment efficiency measurement indicator system after screening based on the LASSO algorithm; The weight calculation module is used to combine the entropy weight method and the hierarchical analysis method to obtain the objective weight and subjective weight of each output indicator respectively. Based on game theory, the two types of weights are combined to calculate the final weight of each output indicator; The output index calculation module is used to calculate the comprehensive output index of each dimension based on the final weight of the output indicator using a linear weighting method; The investment efficiency calculation module is used to calculate the investment efficiency of enterprises based on the super-efficiency SBM model, sort and output them in combination with the calculation structure.

2. The enterprise investment efficiency calculation system based on LASSO-game theory combined weighting-super efficiency SBM according to claim 1 is characterized in that: The expected output indicators include economic benefits and social benefits, and the undesirable output indicators include environmental pollution.

3. The enterprise investment efficiency calculation system based on LASSO-game theory combined weighting-super efficiency SBM according to claim 2 is characterized in that: The environmental pollution indicators include carbon dioxide emissions, sulfur dioxide emissions, and smoke emissions.

4. The enterprise investment efficiency calculation system based on LASSO-game theory combined weighting-super efficiency SBM according to claim 2 is characterized in that: In the data screening module, the LASSO regression method is used to examine the relationship between the investment efficiency evaluation index and corporate performance of all enterprise samples; the details are as follows: Assume that in the evaluation index system construction module, there are R indicators in the enterprise investment efficiency evaluation index system constructed; there are M groups of enterprise samples, and each group of enterprise samples includes R evaluation indicators and corresponding enterprise performance of enterprises within a preset statistical period; Building a regression model where y i is the enterprise performance corresponding to the i-th group of enterprise samples, x ir is the rth evaluation index in the i-th group of enterprise samples, β r is the regression coefficient corresponding to the rth indicator; β0 is the constant term; ε i is the random error term; Optimization objective based on LASSO regression Select the penalty coefficient λ that minimizes the cross-validation error to determine the optimal parameter; The optimal penalty coefficient is used to fit the LASSO regression, the optimal regression coefficient is solved, and the indicators with a regression coefficient of 0 are eliminated to obtain a new indicator system. At this time, the evaluation system contains P input indicators and Q output indicators, among which the expected output indicators include q1 economic benefit indicators, q2 social benefit indicators, and q3 non-expected output indicators.

5. The enterprise investment efficiency calculation system based on LASSO-game theory combined weighting-super efficiency SBM according to claim 1 is characterized in that: The expected output index and the non-expected output index in the selected enterprise investment efficiency measurement index system are processed by data standardization. The original data is X=(x ij ) M×Q (i=1,2,…M;j=1,2,…Q); where x ij represents the j-th output indicator in the i-th group of enterprise samples; the standardized processing formula is:

6. The enterprise investment efficiency calculation system based on LASSO-game theory combined weighting-super efficiency SBM according to claim 1 is characterized in that: In the weight calculation module, the entropy weight method is used to determine the output indicator weights in the investment efficiency measurement index system of the screened enterprise. The calculation process is as follows: First, calculate the information entropy of each output dimension: in 0≤H j ≤1, when p ij = 0, p ij lnp ij =0; Secondly, calculate the entropy weight of the j-th indicator: in, 7. The enterprise investment efficiency calculation system based on LASSO-game theory combined weighting-super efficiency SBM according to claim 1 is characterized in that: In the weight calculation module, the calculation process of the output indicator weights in the enterprise investment efficiency measurement indicator system after screening is determined by using the hierarchical analysis method is as follows: First, construct a discriminant matrix. For the factors in the same layer of the enterprise investment efficiency measurement index system that belong to each factor in the upper layer after screening, use the paired comparison method and the 1-9 comparison scale to construct a paired comparison matrix until the bottom layer; obtain the discriminant matrix A = (a kl ) n×n ,k,l=1,2,…n, where a kl Indicates the importance of factor k relative to factor l. Both k and l represent the factors of the same layer that belong to each factor of the previous layer. n represents the number of factors. a kk =1,a kl =a lk ; Second, the characteristic equation of the discriminant matrix is used to determine the eigenvector, and the eigenvector is normalized to obtain the weight vector W k , determine the maximum eigenvalue Third, calculate the consistency index CI = 0 indicates perfect agreement, and the larger the CI, the less agreement; Fourth, calculate the consistency ratio RI is the random consistency index. When CR is less than 0.1, the matrix is consistent and the discriminant matrix is reasonable. Finally, the relative weight values of each layer are calculated from bottom to top and multiplied to obtain the indicator weight under the AHP method.

8. The enterprise investment efficiency calculation system based on LASSO-game theory combined weighting-super efficiency SBM according to claim 1 is characterized in that: In the weight calculation module, the two types of weights are combined based on game theory to calculate the final weights of each output indicator. The specific process is as follows: Treat the selected output indicators as the player set N in game theory; Calculate the Shapley value of indicator j Where |N| represents the number of elements in the set of indicators, S is any subset that does not contain j, |S| is the number of participants in subset S, v(S) is the value of the cooperative combination S, that is, the sum of the weights of all indicators in subset S under the entropy weight method and the AHP method, v(S∪{j})-v(S) is the marginal contribution of indicator j in subset S, The weight coefficient of all permutations of indicator j when adding it to subset S to ensure fair weighting; Multiply the Shapley value of each indicator by the sum of the weights of its two weighting methods to obtain the combined weight ω of the indicator j C =V(j)×(ω j E +ω j A ), where ω j E and ω j A These are the two sets of original weight vectors provided by the entropy weight method and the AHP method respectively; Then, all indicator weights are normalized to ensure that the sum of the weights is 1, and the final combined weight of the jth indicator can be obtained.

9. The enterprise investment efficiency calculation system based on LASSO-game theory combined weighting-super efficiency SBM according to claim 1 is characterized in that: The final weights obtained by the game theory combined weighting method are used to weighted sum the output indicators of each dimension to obtain the expected output comprehensive index g and the unexpected output comprehensive index h.

10. The enterprise investment efficiency calculation system based on LASSO-game theory combined weighting-super efficiency SBM according to claim 1 is characterized in that: In the investment efficiency calculation module, the super-efficiency SBM model including undesirable output is used to calculate the investment efficiency index of enterprises and then rank them; the super-efficiency SBM model with BCC structure is: Among them, δ represents the efficiency value of the decision-making unit, subscript 0 is the decision-making unit being evaluated, subscript i represents the i-th enterprise sample, f z0 , g0 and h0 represent the input value, expected output value and unexpected output value of the evaluated decision-making unit 0 on the z-th input indicator respectively; P represents the number of input indicators, f zi 、g i and h i represents the value of the input index, expected output index and unexpected output index of enterprise sample i at the zth time, are the redundancy, expected output deficiency and unexpected output superscalar matrices of the z-th input of the decision-making unit, γ i and γ zi represents the weight variable; If the CCR structure super-efficiency SBM model is selected, remove and This condition; For the final efficiency value, if δ>1, it indicates that the enterprise is super-efficient; if δ=1, it indicates that the enterprise is just at the efficient frontier; if δ<1, it indicates that the enterprise needs to improve efficiency.

Images