Income and cost influence factor analysis method based on cloud model and pLASSO

Abstract

The invention discloses an income and cost influence factor analysis method based on a cloud model and pLASSO, and relates to the technical field of data analysis, and the method comprises the steps: building an industry product influence factor set based on a first object influence factor, and generating a preset evaluation grade; constructing a first object influence factor preliminary rating model based on a preset evaluation grade to obtain a first object influence factor preliminary rating; according to the method, a prediction value is obtained through preliminary rating and screening, a first object influence factor analysis model is constructed, and parameters are adjusted to obtain a first object influence factor with a coefficient value not equal to 0. According to the method, deep insights of experts on industry product income and cost key influence factors are integrated into the model; the model is helped to identify variables really related to the target variable more accurately; through comprehensive utilization of the cloud model and the pLASSO model, important factors influencing cost and income are screened out more accurately.

Description

Technical Field

[0001] The present invention relates to the technical field of data analysis, and in particular to a method for analyzing factors affecting revenue and costs based on a cloud model and pLASSO. Background Art

[0002] Understanding the key factors affecting the revenue and costs of typical industry products in the electricity market will help power grid companies deepen their understanding of the trends of trading entities, grasp their development trends and future changes, and provide data support conditions for timely and accurate identification of market risks and the potential for increasing supply and sales.

[0003] Current influencing factor analysis models generally use regression analysis, machine learning and other technical means to find important influencing factors based solely on data. This approach is driven solely by data and ignores the views of domain experts who may provide in-depth insights into key influencing factors of industry product revenue and cost. Such domain knowledge can help the model more accurately identify variables that are truly relevant to the target variable, while making the model's decision-making process more transparent and easy to understand, thereby making more effective decisions.

[0004] When combining expert evaluation and data-driven, the present invention uses the pLASSO model to integrate prior information into LASSO. LASSO is a popular statistical tool that performs variable selection by minimizing a loss function containing the L1 norm. In pLASSO, prior information is used to guide the regularization process of LASSO regression, which improves the predictive ability and interpretability of the model, making the model more in line with the needs of practical applications.

[0005] When converting the quantitative scores of the experts on the importance of the influencing factors into qualitative importance, the present invention uses a cloud model. The cloud model can realize the conversion between qualitative concepts and quantitative descriptions, and is particularly suitable for expressing and processing fuzziness, randomness and incomplete information. Summary of the invention

[0006] In view of the above-mentioned problems, the present invention is proposed.

[0007] Therefore, the technical problem solved by the present invention is: the technical means of the prior art are single, and how to solve the problem that the prior art is only driven by data and ignores the opinions of experts in the field.

[0008] To solve the above technical problems, the present invention provides the following technical solutions: a revenue and cost influencing factor analysis method based on a cloud model and pLASSO, comprising constructing an industry product influencing factor set based on a first object influencing factor, and generating a preset evaluation level; constructing a preliminary rating model for the first object influencing factor based on the preset evaluation level, and obtaining a preliminary rating for the first object influencing factor; obtaining a predicted value based on the preliminary rating screening, constructing a first object influencing factor analysis model, and adjusting parameters to obtain a first object influencing factor whose coefficient value is not 0.

[0009] As a preferred solution of the revenue and cost influencing factor analysis method based on cloud model and pLASSO described in the present invention, the preset evaluation levels include very important, relatively important, general, and unimportant.

[0010] As a preferred solution of the revenue and cost influencing factor analysis method based on cloud model and pLASSO described in the present invention, wherein: the construction of a preliminary rating model for the first object influencing factor based on a preset evaluation grade includes dividing the evaluation grade score interval, and asking experts to score the importance of the first object influencing factor based on experience according to the evaluation grade interval to obtain the evaluation grade of the first object influencing factor; converting the evaluation grades into digital features of the preliminary rating model respectively, and constructing the preliminary rating model.

[0011] As a preferred solution of the revenue and cost influencing factor analysis method based on cloud model and pLASSO described in the present invention, wherein: obtaining the preliminary rating of the first object influencing factor includes calculating the model parameters of the first object influencing factor through the preliminary rating model using the collected expert evaluation data; using the model method to calculate the membership of the influencing factor to the evaluation level respectively; and obtaining the preliminary rating of the first object influencing factor according to the principle of maximum membership.

[0012] As a preferred solution of the revenue and cost influencing factor analysis method based on cloud model and pLASSO described in the present invention, wherein: the prediction value obtained according to the preliminary rating screening includes taking the preliminary rating of the first object influencing factor obtained by the preliminary rating model as prior information, and taking the influencing factors with evaluation levels of "general", "relatively important" and "very important" as the preliminary important influencing factors of the first object; through the screening model, a penalty is imposed on the first object influencing factors rated as "unimportant", so that the important influencing factors of the first object are retained in the preliminary rating model, and the estimated parameters are obtained and the prediction value is calculated.

[0013] As a preferred solution of the revenue and cost influencing factor analysis method based on cloud model and pLASSO described in the present invention, wherein: the construction of the first object influencing factor analysis model includes adding a loss function composed of the predicted value calculated by the preliminary rating model on the basis of the original screening model loss function to construct the first object influencing factor analysis model.

[0014] As a preferred scheme of the income and cost influencing factor analysis method based on cloud model and pLASSO described in the present invention, wherein: the first object influencing factor whose coefficient value is not 0 obtained by adjusting the parameters includes, based on the first object influencing factor analysis model, selecting the adjustment parameters of the analysis model through cross-validation so that the loss function reaches the minimum model coefficient estimate; when the obtained model coefficient estimate is not 0, the obtained first object influencing factor is the screened out important influencing factor of the first object.

[0015] Another object of the present invention is to provide a revenue and cost influencing factor analysis system based on cloud model and pLASSO, which can convert experts' quantitative scores on the importance of influencing factors into qualitative importance through the cloud model, solving the conversion problem between qualitative concepts and quantitative descriptions.

[0016] As a preferred solution of the revenue and cost influencing factor analysis system based on cloud model and pLASSO described in the present invention, it includes: a preset evaluation level module, a preliminary rating module, and a screening module; the preset evaluation level module is used to construct an industry product revenue and cost influencing factor set based on product revenue and cost influencing factors, and generate a preset evaluation level; the preliminary rating module is used to construct a preliminary rating model for revenue and cost influencing factors based on the preset evaluation level, and obtain a preliminary rating of the influencing factors; the screening module is used to obtain a predicted value based on the preliminary rating screening, construct an influencing factor analysis model, and adjust the parameters to obtain influencing factors with coefficient values ​​not being 0.

[0017] A computer device includes a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to implement a method for analyzing factors affecting revenue and costs based on a cloud model and pLASSO.

[0018] A computer-readable storage medium stores a computer program, which, when executed by a processor, implements the steps of a method for analyzing factors affecting revenue and costs based on a cloud model and pLASSO.

[0019] Beneficial effects of the present invention: The revenue and cost influencing factor analysis method based on the cloud model and pLASSO provided by the present invention integrates the expert's in-depth insights into the key influencing factors of industry product revenue and cost into the model, helping the model to more accurately identify the variables that are truly related to the target variable; by using the cloud model to convert quantitative expert scores into qualitative importance ratings, the knowledge and experience of different experts are integrated, the ambiguity and randomness of expert scores are effectively expressed and processed, and the robustness of the model is improved; by integrating expert prior information into the LASSO model, not only the predictive ability and interpretability of the model are improved, but also the model parameters can be automatically adjusted according to the quality of prior information to achieve a balance between estimation efficiency and prediction accuracy; by comprehensively utilizing the cloud model and the pLASSO model, not only the important factors affecting cost and revenue can be screened out more accurately, but also data support can be provided for power grid companies to grasp the trends of trading entities, discover market risks, and increase supply and sales potential. The present invention achieves better results in terms of data multi-party and the balance between estimation efficiency and prediction accuracy. BRIEF DESCRIPTION OF THE DRAWINGS

[0020] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings required for use in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other accompanying drawings can be obtained based on these accompanying drawings without paying creative work.

[0021] Figure 1 An overall flow chart of a revenue and cost influencing factor analysis method based on a cloud model and pLASSO is provided for the first embodiment of the present invention.

[0022] Figure 2 A specific step diagram of a method for analyzing factors affecting revenue and costs based on a cloud model and pLASSO is provided for the second embodiment of the present invention.

[0023] Figure 3 An overall flow chart of a revenue and cost influencing factor analysis system based on a cloud model and pLASSO is provided for the fourth embodiment of the present invention. DETAILED DESCRIPTION

[0024] In order to make the above-mentioned purposes, features and advantages of the present invention more obvious and easy to understand, the specific implementation methods of the present invention are described in detail below in conjunction with the drawings of the specification. Obviously, the described embodiments are part of the embodiments of the present invention, but not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary persons in the art without creative work should fall within the scope of protection of the present invention.

[0025] Example 1, reference Figure 1 , is an embodiment of the present invention, and provides a revenue and cost influencing factor analysis method based on a cloud model and pLASSO, including: constructing an industry product influencing factor set based on a first object influencing factor, and generating a preset evaluation level; constructing a preliminary rating model for the first object influencing factor based on the preset evaluation level, and obtaining a preliminary rating for the first object influencing factor; obtaining a predicted value based on the preliminary rating screening, constructing a first object influencing factor analysis model, and adjusting parameters to obtain a first object influencing factor whose coefficient value is not 0.

[0026] Furthermore, the first object may be product revenue, product cost, product quality, consumer demand, management expenses, etc.

[0027] Furthermore, the method of collecting the first object data and the influencing factor data may be through financial statements, market research, expert interviews, industry data, data analysis software, etc.; financial statements may be obtained by obtaining sales revenue, profit, raw material cost, labor cost, manufacturing cost, management cost and other data from the company's financial statements; market research may be collected through questionnaires, interviews and other methods to collect information such as consumer purchasing behavior, price sensitivity, consumer demand, competitor situation, market trends, etc.; expert interviews may be obtained by interviewing industry experts, scholars, etc. to obtain professional insights into influencing factors; industry data may be obtained by obtaining the overall industry revenue data and publicly available data such as macroeconomic data, population data, consumption data, etc. from industry associations, government agencies and other institutions or departments; data analysis software may be obtained by using Excel, SPSS, python and other data analysis software for data cleaning, analysis and visualization.

[0028] It should be noted that the preset evaluation level can be set to a scoring range of 0 to 100, where the scoring range of "unimportant" is [0, 60], the scoring range of "general" is [61, 75], the scoring range of "relatively important" is [76, 90], and the scoring range of "relatively important" is [91, 100].

[0029] It should also be noted that the preliminary rating is to convert the four evaluation intervals into digital features of the cloud model through the cloud model, and use the collected expert evaluation data through the reverse cloud generator to calculate the cloud parameters of each influencing factor of the first object, and use the forward cloud generator to calculate the membership of each influencing factor of the first object to the evaluation level, so as to obtain the preliminary rating of each influencing factor of the first object.

[0030] It should also be noted that the screening model in the step of screening to obtain the predicted value uses the LASSO model. By adding a loss function to construct pLASSO, adjusting parameters are selected according to the cross-validation method, and the coefficient estimate of pLASSO is obtained by minimizing the loss function to screen out important influencing factors of the first object.

[0031] It should also be noted that the reason for choosing the cloud model and the pLASSO model in the model selection is that the cloud model can effectively solve the conversion problem between qualitative concepts and quantitative descriptions, and better express and handle the uncertainty of expert ratings, thereby improving the accuracy and effectiveness of the first-object influencing factor analysis method; the pLASSO model can effectively solve the problems of variable selection, overfitting, and prior information fusion, thereby improving the accuracy and effectiveness of the first-object influencing factor analysis method.

[0032] Example 2, reference Figure 2 , which is an embodiment of the present invention, provides a method for analyzing factors affecting revenue and costs based on a cloud model and pLASSO, including:

[0033] S1: Construct an industry product influencing factor set based on the first object influencing factor, and generate a preset evaluation level.

[0034] Furthermore, the preset evaluation levels include very important, relatively important, average, and not important.

[0035] It should be noted that if Figure 2 S101 shown in FIG. 1 is to construct a first object influencing factor set based on the first object influencing factor and generate a preset evaluation level. The specific steps are to construct a set of possible influencing factors {x1, x2, x3…x p}, each expert is requested to rate the importance of each influencing factor of the first object according to the evaluation level shown in Table 1. The higher the score, the more important the corresponding influencing factor.

[0036] Table 1 Evaluation level interval table

[0037]

[0038]

[0039] S2: Construct a preliminary rating model for the first object's influencing factors based on the preset evaluation level to obtain a preliminary rating for the first object's influencing factors.

[0040] Furthermore, constructing a preliminary rating model for the first object influencing factors based on preset evaluation levels includes dividing the evaluation level score intervals, asking experts to score the importance of the first object influencing factors based on experience according to the evaluation level intervals, and obtaining the evaluation level of the first object influencing factors; converting the evaluation levels into digital features of the preliminary rating model, and constructing the preliminary rating model.

[0041] like Figure 2 The specific implementation steps of S201, which constructs a preliminary rating model of the first object influencing factors based on the preset evaluation level, are as follows: by converting the four evaluation intervals into digital features of the cloud model, namely, the three parameter expectations of the cloud model Ex m 、Entropy m , variance He m , m = 1, ..., 4, construct the preliminary rating cloud model of the first object influencing factors, where Ex m The calculation formula is expressed as:

[0042]

[0043] En m The calculation formula is expressed as:

[0044]

[0045] He m The calculation formula is expressed as:

[0046] He m =k

[0047] Among them, Ex m 、En m 、He m They represent the expectation, entropy and variance of the mth evaluation interval respectively. represents the minimum value of each evaluation interval, represents the maximum value of each evaluation interval, k represents a constant, and generally takes a value between 0.001 and 0.1. In this embodiment, according to the fuzziness of the comment itself and referring to previous literature, k is set to be 0.1.

[0048] It should be noted that obtaining a preliminary rating of the influencing factors of the first object includes calculating the model parameters of the influencing factors of the first object through the preliminary rating model using the collected expert evaluation data; calculating the membership of the influencing factors to the evaluation level respectively using the model method; and obtaining the preliminary rating of the influencing factors of the first object based on the principle of maximum membership.

[0049] like Figure 2As shown in S202, the specific implementation step of obtaining the preliminary rating of the influencing factors of the first object is to use the collected expert evaluations to calculate the cloud parameters of each influencing factor of the first object using the reverse cloud generator; Figure 2 In S203 shown, the membership of the j-th influencing factor to the m-th evaluation level is calculated using the forward cloud generator; finally, according to the maximum membership principle, the preliminary rating of each influencing factor of the first object is obtained.

[0050] It should also be noted that the specific execution steps of the reverse cloud generator include:

[0051] Input: For the jth influencing factor x j , assuming the sample point is x ij , i=1,...,n;

[0052] Step 1: Calculate expected Ex j , expressed as:

[0053]

[0054] Step 2: Calculate entropy En j , expressed as:

[0055]

[0056] Step 3: Calculate the variance He j , expressed as:

[0057]

[0058] Output: Numerical features Ex reflecting qualitative concepts j 、En j 、He j ;

[0059] It should also be noted that the specific execution steps of the forward cloud generator include:

[0060] Input: The numerical feature Ex of the jth influencing factor j 、En j 、He j , the expectation Ex of the mth evaluation level m 、Entropy m , simulation times N;

[0061] Step 1: Generate En j is the expected value, He j 2 A normal random number y with variance i ;

[0062] Step 2: Generate Ex j is the expected value, yi 2 A normal random number z with variance i ;

[0063] Step 3: Calculate the membership of the i-th simulation It is expressed as:

[0064]

[0065] Step 4: Repeat steps 1 to 3 N times to calculate the mean μ of N membership degrees jm , expressed as:

[0066]

[0067] Output: The membership degree μ of the jth influencing factor to the mth evaluation level jm ;

[0068] According to the principle of maximum membership, a preliminary rating of each influencing factor of the first object is obtained.

[0069] S3: The predicted value is obtained according to the preliminary rating screening, and the first object influencing factor analysis model is constructed, and the parameters are adjusted to obtain the first object influencing factor whose coefficient value is not 0.

[0070] Furthermore, obtaining the predicted value based on the preliminary rating screening includes taking the preliminary rating of the first object influencing factors obtained by the preliminary rating model as prior information, and taking the influencing factors with evaluation levels of "general", "relatively important" and "very important" as preliminary important influencing factors of the first object; through the screening model, imposing a penalty on the first object influencing factors rated as "unimportant", so that the important influencing factors of the first object are retained in the preliminary rating model, and the estimated parameters are obtained and the predicted value is calculated.

[0071] like Figure 2 As shown in S301, the specific implementation step of obtaining the predicted value according to the preliminary rating screening is to use the first object influencing factor rating obtained based on the cloud model as prior information, and use the influencing factors rated as "general", "relatively important" and "very important" as the first object preliminary important influencing factors, assuming that S = {x1, x2, x3...x k} are the important influencing factors obtained through the cloud model.

[0072] The collected first object influencing factor data and the first object data are standardized, and the LASSO model is used to impose penalties on the influencing factors that are initially rated as "unimportant", so that all the influencing factors in S are retained in the model, and the LASSO model loss function is constructed, which is expressed as:

[0073]

[0074] Among them, x qj represents the value of the jth influencing factor of the qth enterprise, q=1,...,N, j=1,...,p, Y=(Y1,...,Y N ) T , Y q represents the revenue and cost of the qth enterprise, represents the coefficient to be estimated, and γ represents the adjustment parameter;

[0075] The adjustment parameter γ is selected according to the cross-validation method, and the estimated parameter is obtained by calculation. It is expressed as:

[0076]

[0077] Calculate predicted values It is expressed as:

[0078]

[0079] in, Represents the revenue and cost forecast values ​​of the qth enterprise.

[0080] It should be noted that constructing the first object influencing factor analysis model includes adding a loss function consisting of the predicted values ​​calculated by the preliminary rating model on the basis of the original screening model loss function to construct the first object influencing factor analysis model.

[0081] like Figure 2 The specific implementation step of constructing the first object influencing factor analysis model in S302 is as follows: Construct the pLASSO model by adding the predicted value to the LASSO model loss function. The loss function constructed;

[0082] The loss function of the pLASSO model is constructed as:

[0083]

[0084] in, It represents the secondary loss calculated using the original income and cost data Y;

[0085] in, Indicates forecasted revenue and cost data Calculated quadratic loss;

[0086] Where β=(β1,...,β p ) T represents the coefficient to be estimated, and λ and η represent the adjustment parameters.

[0087] It should also be noted that the first object influencing factor whose coefficient value is not 0 obtained by adjusting the parameters includes, based on the first object influencing factor analysis model, selecting the adjustment parameters of the analysis model through cross-validation so that the loss function reaches the minimum model coefficient estimate; when the obtained model coefficient estimate is not 0, the obtained first object influencing factor is the screened out important influencing factor of the first object.

[0088] like Figure 2 The specific implementation steps of adjusting the parameters to obtain the first object influencing factor whose coefficient value is not 0 are as follows: selecting the adjustment parameters λ, η according to the cross-validation method; and obtaining the coefficient estimation value of pLASSO by minimizing the loss function of the pLASSO model. The first object influencing factors whose coefficient values ​​are not 0 are the screened first object important influencing factors.

[0089] Example 3, an embodiment of the present invention, provides a method for analyzing factors affecting income and costs based on a cloud model and pLASSO. In order to verify the beneficial effects of the present invention, scientific demonstration is carried out through economic benefit calculation and simulation experiments.

[0090] First, a set of possible influencing factors affecting the cost of the electrolytic aluminum industry was constructed, resulting in a total of 27 variables as shown in Table 1 below.

[0091] Table 1 A collection of possible factors affecting the cost of electrolytic aluminum industry

[0092]

[0093]

[0094] Secondly, the importance scores of the above 27 variables on the production cost of electrolytic aluminum enterprises were collected from 50 experts. According to the method of the present invention, the importance of influencing factors was preliminarily evaluated using the cloud model, and 12 preliminary important influencing factors were obtained, as shown in Table 2 below.

[0095] Table 2 Preliminary important influencing factors obtained using the cloud model

[0096] Labor cost per ton of output Asset turnover Proportion of automated equipment Logistics and transportation costs per ton of production R&D output rate Inventory Turnover Raw material cost per ton of output Return on R&D investment Management expenses ratio Energy efficiency Gross Domestic Product Operating profit margin

[0097] Finally, data on the above 27 possible influencing factors for 100 enterprises in the electrolytic aluminum industry during a certain period of time and the total cost data required for each ton of output of the 100 enterprises during this period are collected as Y.

[0098] According to the method of the present invention, all data are first standardized, and the cost forecast value of each enterprise is obtained using the LASSO model. The pLASSO model was constructed, and the regression coefficients of each influencing factor were finally obtained as shown in Table 3.

[0099] Table 3 Regression coefficients of various influencing factors obtained by pLASSO model

[0100]

[0101]

[0102] The influencing factors whose coefficient values ​​are not 0 are the important influencing factors screened out.

[0103] It can be seen from the screening results that this embodiment screened out a total of 15 important influencing factors, which have both overlaps and deviations with the preliminary important factors obtained by expert scoring through the cloud model. Therefore, it can be seen that the result combines expert scoring and actual data, and provides more detailed insights and analysis compared to using only the LASSO model for variable selection.

[0104] Example 4, reference Figure 3 , as an embodiment of the present invention, provides a revenue and cost influencing factor analysis system based on cloud model and pLASSO, including a preset evaluation level module, a preliminary rating module, and a screening module.

[0105] Among them, the preset evaluation level module is used to construct a set of industry product revenue and cost influencing factors based on product revenue and cost influencing factors, and generate preset evaluation levels; the preliminary rating module is used to construct a preliminary rating model for revenue and cost influencing factors based on the preset evaluation levels, and obtain preliminary ratings of influencing factors; the screening module is used to obtain predicted values ​​based on preliminary rating screening, construct an influencing factor analysis model, and adjust parameters to obtain influencing factors whose coefficient values ​​are not 0.

[0106] If the function is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, or the part that contributes to the prior art, or the part of the technical solution, can be embodied in the form of a software product. The computer software product is stored in a storage medium, including several instructions for a computer device (which can be a personal computer, server, or network device, etc.) to perform all or part of the steps of the methods of each embodiment of the present invention. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), disk or optical disk, etc. Various media that can store program codes.

[0107] The logic and / or steps represented in the flowchart or otherwise described herein, for example, can be considered as an ordered list of executable instructions for implementing logical functions, and can be embodied in any computer-readable medium for use by an instruction execution system, device or apparatus (such as a computer-based system, a system including a processor, or other system that can fetch instructions from an instruction execution system, device or apparatus and execute instructions), or in conjunction with such instruction execution systems, devices or apparatuses. For the purposes of this specification, "computer-readable medium" can be any device that can contain, store, communicate, propagate or transmit a program for use by an instruction execution system, device or apparatus, or in conjunction with such instruction execution systems, devices or apparatuses.

[0108] More specific examples of computer-readable media (a non-exhaustive list) include the following: an electrical connection with one or more wires (electronic device), a portable computer disk case (magnetic device), a random access memory (RAM), a read-only memory (ROM), an erasable and programmable read-only memory (EPROM or flash memory), an optical fiber device, and a portable compact disk read-only memory (CDROM). In addition, the computer-readable medium may even be a paper or other suitable medium on which the program is printed, since the program may be obtained electronically, for example, by optically scanning the paper or other medium, followed by editing, deciphering, or processing in another suitable manner as necessary, and then stored in a computer memory.

[0109] It should be understood that the various parts of the present invention can be implemented by hardware, software, firmware or a combination thereof. In the above-mentioned embodiments, multiple steps or methods can be implemented by software or firmware stored in a memory and executed by a suitable instruction execution system. For example, if implemented by hardware, as in another embodiment, it can be implemented by any one of the following technologies known in the art or their combination: a discrete logic circuit having a logic gate circuit for implementing a logic function for a data signal, a dedicated integrated circuit having a suitable combination of logic gate circuits, a programmable gate array (PGA), a field programmable gate array (FPGA), etc. It should be noted that the above embodiments are only used to illustrate the technical solution of the present invention and are not limited. Although the present invention is described in detail with reference to the preferred embodiments, it should be understood by those skilled in the art that the technical solution of the present invention can be modified or replaced by equivalents without departing from the spirit and scope of the technical solution of the present invention, which should be included in the scope of the claims of the present invention.

[0110] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solutions of the present invention may be modified or replaced by equivalents without departing from the spirit and scope of the technical solutions of the present invention, which should all be included in the scope of the claims of the present invention.

Claims

1. A method for analyzing factors affecting revenue and costs based on cloud model and pLASSO, characterized in that: include: Constructing a set of industry product influencing factors based on the first object influencing factors, and generating a preset evaluation level; Constructing a preliminary rating model for the influencing factors of the first object based on the preset evaluation level, and obtaining a preliminary rating for the influencing factors of the first object; The predicted value is obtained according to the preliminary rating screening, the first object influencing factor analysis model is constructed, and the parameters are adjusted to obtain the first object influencing factor with a coefficient value not being 0.

2. The method for analyzing factors affecting income and costs based on cloud model and pLASSO as claimed in claim 1, characterized in that: The preset evaluation levels include very important, relatively important, general, and unimportant.

3. The method for analyzing factors affecting income and cost based on cloud model and pLASSO as claimed in claim 2, characterized in that: The constructing of a preliminary rating model for the first object influencing factor based on the preset evaluation grade includes dividing the evaluation grade score intervals, and asking experts to score the importance of the first object influencing factor based on experience according to the evaluation grade intervals to obtain the evaluation grade of the first object influencing factor; The evaluation levels are converted into digital features of the preliminary rating model respectively, and the preliminary rating model is constructed.

4. The method for analyzing factors affecting income and cost based on cloud model and pLASSO as claimed in claim 3, characterized in that: Obtaining a preliminary rating of the influencing factors of the first object includes calculating model parameters of the influencing factors of the first object through a preliminary rating model using the collected expert evaluation data; Using the model method, the membership of the influencing factors to the evaluation level is calculated respectively; According to the principle of maximum membership, a preliminary rating of the influencing factors of the first object is obtained.

5. The method for analyzing factors affecting income and cost based on cloud model and pLASSO as claimed in claim 4, characterized in that: The obtaining of the prediction value according to the preliminary rating screening includes taking the preliminary rating of the influencing factors of the first object obtained by the preliminary rating model as prior information, and taking the influencing factors with evaluation levels of "general", "relatively important" and "very important" as preliminary important influencing factors of the first object; Through the screening model, penalties are imposed on the first object influencing factors rated as "unimportant", so that all important influencing factors of the first object are retained in the preliminary rating model, and the estimated parameters are obtained and the predicted values ​​are calculated.

6. The method for analyzing factors affecting income and costs based on cloud model and pLASSO as claimed in claim 5, characterized in that: The construction of the first object influencing factor analysis model includes adding a loss function consisting of predicted values ​​calculated by the preliminary rating model on the basis of the original screening model loss function to construct the first object influencing factor analysis model.

7. The method for analyzing factors affecting income and cost based on cloud model and pLASSO as claimed in claim 6, characterized in that: The adjusting parameter to obtain the first object influencing factor whose coefficient value is not 0 includes selecting the adjusting parameter of the analysis model through cross-validation based on the first object influencing factor analysis model so that the loss function reaches the minimum model coefficient estimation value; When the obtained estimated value of the model coefficient is not 0, the obtained first object influencing factor is the screened out important influencing factor of the first object.

8. A system using the revenue and cost influencing factor analysis method based on cloud model and pLASSO as claimed in any one of claims 1 to 7, characterized in that: It includes preset evaluation level module, preliminary rating module and screening module; The preset evaluation grade module is used to construct an industry product influencing factor set based on the first object influencing factor and generate a preset evaluation grade; The preliminary rating module is used to construct a preliminary rating model of the first object influencing factors based on the preset evaluation level to obtain a preliminary rating of the first object influencing factors; The screening module is used to obtain a prediction value according to the preliminary rating screening, construct a first object influencing factor analysis model, and adjust parameters to obtain a first object influencing factor whose coefficient value is not 0.

9. A computer device comprising a memory and a processor, wherein the memory stores a computer program, wherein: When the processor executes the computer program, the steps of the revenue and cost influencing factor analysis method based on the cloud model and pLASSO are implemented as described in any one of claims 1 to 7.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the revenue and cost influencing factor analysis method based on a cloud model and pLASSO are implemented as described in any one of claims 1 to 7.

Images