CN112381653A - Mining and recommending method, device, equipment and storage medium for potential investment enterprises - Google Patents

Mining and recommending method, device, equipment and storage medium for potential investment enterprises Download PDF

Info

Publication number
CN112381653A
CN112381653A CN202011177090.3A CN202011177090A CN112381653A CN 112381653 A CN112381653 A CN 112381653A CN 202011177090 A CN202011177090 A CN 202011177090A CN 112381653 A CN112381653 A CN 112381653A
Authority
CN
China
Prior art keywords
target
investment
enterprise
label
model
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Pending
Application number
CN202011177090.3A
Other languages
Chinese (zh)
Inventor
郭亚萌
张炫
龙非池
贺光灿
刘帅
何浪
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Chongqing Shake Big Data Co ltd
Original Assignee
Chongqing Shake Big Data Co ltd
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Chongqing Shake Big Data Co ltd filed Critical Chongqing Shake Big Data Co ltd
Priority to CN202011177090.3A priority Critical patent/CN112381653A/en
Publication of CN112381653A publication Critical patent/CN112381653A/en
Pending legal-status Critical Current

Links

Images

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q40/00—Finance; Insurance; Tax strategies; Processing of corporate or income taxes
    • G06Q40/06—Asset management; Financial planning or analysis
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
    • G06F16/90—Details of database functions independent of the retrieved data types
    • G06F16/95—Retrieval from the web
    • G06F16/953—Querying, e.g. by the use of web search engines
    • G06F16/9535—Search customisation based on user profiles and personalisation
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q10/00—Administration; Management
    • G06Q10/04—Forecasting or optimisation specially adapted for administrative or management purposes, e.g. linear programming or "cutting stock problem"
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q10/00—Administration; Management
    • G06Q10/06—Resources, workflows, human or project management; Enterprise or organisation planning; Enterprise or organisation modelling
    • G06Q10/067—Enterprise or organisation modelling
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F2216/00—Indexing scheme relating to additional aspects of information retrieval not explicitly covered by G06F16/00 and subgroups
    • G06F2216/03—Data mining

Landscapes

  • Engineering & Computer Science (AREA)
  • Business, Economics & Management (AREA)
  • Human Resources & Organizations (AREA)
  • Strategic Management (AREA)
  • Theoretical Computer Science (AREA)
  • Economics (AREA)
  • Development Economics (AREA)
  • Entrepreneurship & Innovation (AREA)
  • General Physics & Mathematics (AREA)
  • Physics & Mathematics (AREA)
  • Game Theory and Decision Science (AREA)
  • Operations Research (AREA)
  • Marketing (AREA)
  • Databases & Information Systems (AREA)
  • General Business, Economics & Management (AREA)
  • Quality & Reliability (AREA)
  • Finance (AREA)
  • Tourism & Hospitality (AREA)
  • Accounting & Taxation (AREA)
  • Technology Law (AREA)
  • Educational Administration (AREA)
  • Data Mining & Analysis (AREA)
  • General Engineering & Computer Science (AREA)
  • Management, Administration, Business Operations System, And Electronic Commerce (AREA)
  • Information Retrieval, Db Structures And Fs Structures Therefor (AREA)

Abstract

The invention provides a mining and recommending method, device, equipment and storage medium for potential investment enterprises, wherein the method comprises the following steps: designing a general investment willingness label of an enterprise according to a preset potential label value formula; establishing an initial submerged excavation model based on a limit gradient lifting algorithm according to the investment wish label; acquiring historical success data, and training an initial submerged excavation model to obtain a target submerged excavation model; acquiring characteristics of a target enterprise, wherein the characteristics are used for representing basic data of the target enterprise; forecasting the investment willingness of the target enterprise according to the characteristics and the target latent investment mining model to obtain a forecasted target willingness value; and recommending the target enterprises with the target willingness values higher than the preset threshold value. The scheme measures the investment potential of an enterprise in an index mode, realizes the mining of potential investors, and realizes the recommendation of the potential investors, thereby reducing the labor and time cost and improving the conversion efficiency of the potential investors.

Description

Mining and recommending method, device, equipment and storage medium for potential investment enterprises
Technical Field
The invention relates to the technical field of computer communication, in particular to a mining and recommending method, device, equipment and storage medium for potential investment enterprises.
Background
In the face of complicated and changeable international and domestic economy, accurate recruiters are inevitable choices for improving regional economy core competitiveness and adapting to activity regularity of the recruiters. How to improve the recruitment work of the people, implement the accurate recruitment of big data and optimize the industrial structure is the key point of research at present. The most important is the accurate recruitment problem, how to select the optimal enterprise through a big data technology, and the enterprise which is most suitable and has investment will take the best care of the enterprise, but because of the influence of factors such as performance and information asymmetry, blind recruitment in various places can be caused, and the degree of agreeing between the accuracy of the recruitment and the industry is not considered completely. Finally, the business recruiter can not drive the development of the local economy, even needs government support, and the development of the local economy is hindered. Particularly in most regions with relatively delayed economic development, the business investment is still the point in government economic work, and is an important hand for promoting economic development. However, with new changes of economic development modes and global economic development environments in China, the recruitment and investment introduction work also faces brand new problems and challenges.
At present, the business inviting forms mainly adopted in various places are largely the same and slightly different. The main body of the recruiter can be divided into government recruiters (including all levels of governments, economic development parks and the like), intermediary recruiters (including various business associations, professional recruiter organizations and the like), commercial recruiters and the like; according to the environment of the recruiter, the method can be divided into home recruiter, residence recruiter, conference (recommendation) recruiter and the like; according to the content of the recruiter, the industry chain recruiter, the resource (including the resources for inventory), the special recruiter and the like can be divided. These recruiters have various forms, various contents and various effects, and have wide acceptance and application, but have common problems that the analysis depth of resources is limited. The recruitment of funds requires solving the problem of maximizing the resource-to-project matching effect. Therefore, jumping out of a small space and putting resources into a large environment, qualitative and quantitative analysis of the influence factors of various resources become a prerequisite for effective matching of resources. In this respect, the traditional manner of soliciting merchants is far from adequate. Secondly, the direction of introducing the target project is not clear, and the potential is not deeply explored. The resource usage of the project is of opportunity cost, and whether the project is most suitable for the provided resource needs to be comprehensively evaluated from the aspects of resource endowment, project development potential and the like. Particularly, considering the continuous change of the current economic development mode and enterprise development mode, and the fact that the appraisers are usually based on self experience to subjectively evaluate the project, the sustainability and the fitness of the appraisal project are difficult to grasp. Third, the limited number and experience of the recruiters makes them have a limited amount of resources available, resulting in a decentralised recruitment effort and poor results.
Disclosure of Invention
In view of the above, there is a need to provide a mining and recommending method, device, apparatus and storage medium for potential investment enterprises.
A method for mining and recommending potential investment enterprises, the method comprising: designing an investment intention label common to enterprises according to the constituent factors of the investment intention label, and setting a potential label value formula; establishing an initial submerged excavation model based on a limit gradient lifting algorithm according to the investment wish label; obtaining historical success data, and training the initial submerged entry mining model to obtain a target submerged entry mining model; acquiring characteristics of a target enterprise, wherein the characteristics are used for representing basic data of the target enterprise; predicting the investment willingness of the target enterprise according to the characteristics and the target latent investment mining model to obtain a predicted target willingness value; and recommending the target enterprises with the target willingness values higher than the preset threshold value.
In one embodiment, the preset potential tag value formula is specifically:
label＝wDEALDEAL+wINCORRINCORR+wSORELASORELA+wCIINTERCIINTER +wVENDINCVENDINC+wARGARG+wPROGROPROGRO +wRARGRORATGRO+wASSGROASSGRO+wINNOVAINNOVA +wINDEXPINDEXP+wCOMEXPCOMEXP
in one embodiment, the label is a potential label, and DEAL indicates whether the parking has been successful, which is a discrete value, and the successful parking is recorded as 1, otherwise, it is recorded as 0; INCORR represents enterprise investment relevance, which is also a discrete value, and the relation is marked as 1, otherwise, the relation is marked as 0; the SORELA represents the social relevance degree, is a discrete value and has a relevance of 1, otherwise, the social relevance degree is 0; CIINTER represents the enterprise industry association degree, and has an association of 1, otherwise, 0; VENDINC, ARG, PROGRO, ratgo, assro, inova, INDEXP, and COMEXP respectively represent normalized scores and value ranges (0, 1) of an enterprise in terms of "main revenue", "annual ring ratio growth", "profit sum", "tax sum", "asset sum", "innovation capacity", "industry expandability", and "enterprise expandability", respectively; where wi (i ═ VENDINC, ARG, PROGRO, ratgo, assro, inova, INDEXP, COMEXP) respectively represent linear weights of the above indexes in order, and Σ wi ═ 1 is satisfied.
In one embodiment, the label satisfies 0 ≦ label ≦ 1.
The utility model provides a mining and recommendation device of potential investment enterprise, includes label design module, model establishment module, model training module, characteristic acquisition module, will prediction module and enterprise recommendation module, wherein: the label design module is used for designing an investment intention label universal for an enterprise according to the constituent factors of the investment intention label and setting a potential label value formula; the model establishing module is used for establishing an initial submerged excavation model based on a limit gradient lifting algorithm according to the investment willingness label; the model training module is used for acquiring historical success data and training the initial submerged excavation model to obtain a target submerged excavation model; the characteristic acquisition module is used for acquiring the characteristics of the target enterprise, and the characteristics are used for representing basic data of the target enterprise; the intention prediction module is used for predicting the investment intention of the target enterprise according to the characteristics and the target latent investment mining model to obtain a predicted target intention value; the enterprise recommending module is used for recommending the target enterprises with the target willingness values higher than the preset threshold value.
An apparatus comprising a memory, a processor and a computer program stored on the memory and executable on the processor, the processor implementing the steps of the method for mining and recommending potential investment enterprises as described in the various embodiments above when executing the program.
A storage medium having stored thereon a computer program which, when being executed by a processor, carries out the steps of the method for mining and recommending potential investment enterprises as described in the various embodiments above.
According to the mining and recommending method, device, equipment and storage medium for the potential investment enterprises, an initial potential investment mining model is established by designing an investment wish label common to the enterprises and based on a limit gradient lifting algorithm, the model is trained by using historical success data to obtain a target potential investment mining model, target enterprise characteristics are obtained, and a predicted target wish value is obtained based on the target potential investment mining model. The scheme realizes the excavation of potential investors and the recommendation of the potential investors by proposing a potential investment index, namely a label (investment intention), and measuring the investment potential of an enterprise in an indexing way, and gets rid of the dependence of manual judgment in the past, thereby greatly reducing the labor and time cost and greatly improving the conversion efficiency of the potential investors.
Drawings
FIG. 1 is a schematic flow chart diagram illustrating a method for mining and recommending potential investment enterprises in one embodiment;
FIG. 2 is a block diagram of an exemplary mining and recommending apparatus for potential investment enterprises;
fig. 3 is an internal structural diagram of the device in one embodiment.
Detailed Description
In order to make the objects, technical solutions and advantages of the present invention more apparent, the present invention is described in further detail below with reference to the accompanying drawings by way of specific embodiments. It should be understood that the specific embodiments described herein are merely illustrative of the invention and are not intended to limit the invention.
The core of the potential investor is evaluated by the 'investment will' of different enterprises, the mining and recommending method of the potential investor provided by the application is applied to the accurate investment, and is based on a big data method, information of historical successful investment of a target area is used as a training set of a model in the method, so that a target latent investment mining model conforming to the target area is obtained, the target latent investment mining model is used for 'learning' a mapping relation between enterprise characteristics and quantization potential by quantifying the 'investment will' of the enterprises, and the 'enterprise investment desire evaluation problem' is abstracted into a 'multiple regression problem'.
In one embodiment, as shown in fig. 1, there is provided a mining and recommending method for potential investment enterprises, comprising the following steps:
s110, designing an investment intention label common to enterprises according to the composition factors of the investment intention label, and setting a potential label value formula.
Specifically, the investment intention label is "enterprise investment intention" and will be quantized to a continuous value. The composition of the investment intention label is composed of subjective and objective factors and the like, so that an investment intention label common to any enterprise needs to be designed, and the concrete expression is to set a potential label value formula.
In one embodiment, the potential tag value formula in step S110 is specifically:
label＝wDEALDEAL+wINCORRINCORR+wSORELASORELA+wCIINTERCIINTER +wVENDINCVENDINC+wARGARG+wPROGROPROGRO +wRARGRORATGRO+wASSGROASSGRO+wINNOVAINNOVA +wINDEXPINDEXP+wCOMEXPCOMEXP
specifically, the value represented by the potential label value formula is the investment potential of the enterprise, label in the formula is an investment willingness label, variable DEAL represents whether successful parking is performed, a discrete value per se is recorded as 1, and otherwise, the discrete value is recorded as 0; the variable INCORR represents the enterprise investment relevance and is also a discrete value, if the variable INCORR is related, the variable INCORR is marked as 1, otherwise, the variable INCORR is marked as 0; the variable SORELA represents the social relevance degree, is a discrete value, has a relevance of 1, and is 0 otherwise; the variable CIINTER represents the enterprise industry association degree, the association is marked as 1, otherwise, the association is marked as 0; variables are as follows: VENDINC, ARG, PROGRO, ratgo, assro, inova, INDEXP, and COMEXP represent normalized scores and value ranges (0, 1) of an enterprise in terms of "major revenue", "annual ring ratio growth", "profit sum", "tax sum", "asset sum", "innovation capacity", "industry expandability", and "enterprise expandability", respectively.
In one embodiment, label is the willingness to invest label, DEAL indicates whether successful parking has been done, it is a discrete value, successful parking is recorded as 1, otherwise, it is recorded as 0; INCORR represents enterprise investment relevance, which is also a discrete value, and the relation is marked as 1, otherwise, the relation is marked as 0; the SORELA represents the social relevance degree, is a discrete value and has a relevance of 1, otherwise, the social relevance degree is 0; CIINTER represents the enterprise industry association degree, and has an association of 1, otherwise, 0; VENDINC, ARG, PROGRO, ratgo, assro, inova, INDEXP, and COMEXP respectively represent normalized scores and value ranges (0, 1) of an enterprise in terms of "main revenue", "annual ring ratio growth", "profit sum", "tax sum", "asset sum", "innovation capacity", "industry expandability", and "enterprise expandability", respectively; wherein wi(i ═ VENDING, ARG, PROGRO, RATGRO, ASSGRO, INNOVA, INDEXP, COMEXP) respectively represent linear weights of the above-mentioned indexes in order, and satisfy Σ wi＝1。
In one embodiment, label satisfies 0 ≦ label ≦ 1.
Specifically, the closer the label is to 1, the greater the investment potential of the enterprise on the target area.
S120, establishing an initial submerged-projection mining model based on a limit gradient lifting algorithm according to the investment wish label.
S130, historical success data are obtained, the initial submerged excavation model is trained, and the target submerged excavation model is obtained.
Specifically, an algorithm framework used by the latent excavation model is eXtreme Gradient Boosting (XGBoost).
(1)Booster
Booster is a basic model in the XGboost algorithm framework and is a binary regression tree. The training process for the Booster composition-based integration model can be described as: firstly, establishing a first Booster, then gradually iterating, adding one Booster in each iteration until all the boosters are trained, and finally training out a strong evaluator integrating a plurality of boosters. The output of the integrated model is the summation of the leaf node weights output by all booters, namely the potential value of the entrance of one enterprise into the park. The feature is to input the weights of a plurality of leaf nodes into all Booster, and then all the weights are accumulated to be used as the prediction output of the integrated model algorithm.
For an integration model consisting of K boosters, enterprise user characteristics xiThe predicted values obtained by the model are:
Figure BDA0002748995830000061
(2) objective function
In the XGboost framework, an objective function is composed of a loss function and a regular term, and for a training set with N enterprise sample characteristics, the training set is a historical success set of a target region, and the objective function is shown as a formula (1-1):
Figure BDA0002748995830000062
(3) extreme gradient boost
1) Simplifying objective functions
For the t (0 < t ≦ K) iteration, the prediction result of the enterprise customer feature xi in the current model may be described as:
Figure BDA0002748995830000063
a) substituting the formula (1-3) into the objective function (1-2), and simultaneously extracting the t-th term in the regular terms separately, the following can be obtained:
Figure BDA0002748995830000064
b) the equations (1-4) can be Taylor expanded:
Figure BDA0002748995830000065
c) the additive constant term and the higher order infinity term are truncated and let:
Figure BDA0002748995830000066
the objective function (1-5) can be simplified as:
Figure BDA0002748995830000071
d) and substituting the XGboost algorithm into the definitional expression of the regular term. Meanwhile, the form of superposition according to the training sample index is changed into the form of superposition according to the leaf nodes, and the loss function item can be obtained:
Figure 1
2) optimal leaf node weight w
a) The objective function (1-8) is partial-derivative with respect to each leaf node weight and made 0.
Figure BDA0002748995830000073
b) Solving equations (1-9) yields:
Figure BDA0002748995830000074
c) substituting the optimal leaf node weight into a formula (1-8) to obtain an optimal Booster structure score as follows:
Figure BDA0002748995830000075
3) optimal leaf node number T
a) The XGboost adopts a greedy algorithm strategy to realize node splitting judgment, for a leaf node, if the node is not split, T is 1, and according to (1-11), the structure fraction/objective function value of a single leaf node is as follows:
Figure BDA0002748995830000076
b) taking the leaf node as a new root node, and generating two new leaf nodes from the root node according to a certain decision boundary of a certain characteristic in the data set, wherein the sum of the structural scores of the two new leaf nodes is as follows:
Figure BDA0002748995830000081
c) defining checkpoint split structure gains
Figure BDA0002748995830000082
If Gain is greater than 0, the structural scores calculated by the two new nodes are lower than the structural score of the original single node, and the root node is divided into two new leaf nodes according to the optimal judgment boundary of the current characteristics; otherwise no splitting action is performed.
(4) XGboost modeling process
1) Inputting a hyper-parameter.
2) Judging whether K boosters exist
3) And if so, outputting the latent projection evaluation model.
4) If not, establishing a binary lifting tree, then judging whether the maximum depth is reached, searching for a segmentation point, creating left and right leaf nodes based on the segmentation point, distributing samples to the newly created leaf nodes based on the segmentation point, initializing a new leaf node list to be segmented, and calculating w, Gain, G and H with the segmented leaf nodes. And repeating the steps until the steps are completed.
5) And when the 4) is realized, saving the model, wherein the Booster number is +1, and then returning to 2) until the latent projection evaluation model is output.
The implementation steps are as follows: from the beginning of building the first tree, the overall ensemble learning model performance is better than before every time a new tree model is trained. Since the objective function optimization objective is to solve the optimal tree model structure, when the optimization is completed, each sample is divided into the appropriate leaf nodes in each Booster.
S140, characteristics of the target enterprise are obtained, and the characteristics are used for representing basic data of the target enterprise.
Specifically, characteristics of the target enterprise are obtained, and the characteristics are used for representing basic data of the target enterprise; the data characteristics of the enterprises are continuous, such as earnings, profits, taxes, patent numbers and the like of different years; there are also discrete types such as social relevance of investments, industry relevance of enterprises, etc. On the basis of fully utilizing existing data, some new characteristics are constructed, such as: the ring ratio is increased according to the revenue calculation of the enterprise in different years, the development and expansion capability of the whole industry and the expansion capability of the enterprise, namely external investment, concurrent purchase and the like.
S150, according to the characteristics and the target potential investment mining model, the investment intention of the target enterprise is predicted, and a predicted target intention value is obtained.
Specifically, according to the characteristics of the target enterprise and the obtained target latent investment mining model, the investment intention of the target enterprise is predicted, and then the predicted target intention value of the target enterprise can be obtained.
And S160, recommending the target enterprises with the target intention values higher than the preset threshold value.
Specifically, a threshold value is preset, and a target enterprise corresponding to a target will higher than the threshold value is a target enterprise with a possibly higher investment will, so that the enterprise is worthy of being recommended to a place needing to be recruited.
In the embodiment, an initial latent projection mining model is established by designing a general investment intention label of an enterprise and based on a limit gradient lifting algorithm, the model is trained by using historical success data to obtain a target latent projection mining model, target enterprise characteristics are obtained, and a predicted target intention value is obtained based on the target latent projection mining model. The scheme realizes the excavation of potential investors and the recommendation of the potential investors by proposing a potential investment index, namely a label (investment intention), and measuring the investment potential of an enterprise in an indexing way, and gets rid of the dependence of manual judgment in the past, thereby greatly reducing the labor and time cost and greatly improving the conversion efficiency of the potential investors.
In one embodiment, as shown in fig. 2, there is provided a mining and recommending apparatus 200 for potential investment enterprises, which includes a tag designing module 210, a model building module 220, a model training module 230, a feature obtaining module 240, a willingness predicting module 250 and an enterprise recommending module 260, wherein:
the label design module 210 is used for designing an investment intention label common to enterprises according to the constituent factors of the investment intention label and setting a potential label value formula;
the model establishing module 220 is used for establishing an initial submerged excavation model based on a limit gradient lifting algorithm according to the investment wish label;
the model training module 230 is configured to obtain historical success data, train the initial submerged excavation model, and obtain a target submerged excavation model;
the characteristic obtaining module 240 is configured to obtain characteristics of the target enterprise, where the characteristics are used to represent basic data of the target enterprise;
the intention forecasting module 250 is used for forecasting the investment intention of the target enterprise according to the characteristics and the target potential investment mining model to obtain a forecasted target intention value;
the enterprise recommending module 260 is configured to recommend a target enterprise of which the target will value is higher than a preset threshold value.
In one embodiment, a device is provided, which may be a server, and its internal structure diagram may be as shown in fig. 3. The device includes a processor, a memory, a network interface, and a database connected by a system bus. Wherein the processor of the device is configured to provide computing and control capabilities. The memory of the device comprises a nonvolatile storage medium and an internal memory. The non-volatile storage medium stores an operating system, a computer program, and a database. The internal memory provides an environment for the operation of an operating system and computer programs in the non-volatile storage medium. The database of the device is used for storing configuration templates and also can be used for storing target webpage data. The network interface of the device is used for communicating with an external terminal through a network connection. The computer program is executed by a processor to implement a method for mining and recommending potential investment enterprises.
Those skilled in the art will appreciate that the configuration shown in fig. 3 is a block diagram of only a portion of the configuration associated with the present application and does not constitute a limitation on the devices to which the present application may be applied, and that a particular device may include more or less components than those shown, or may combine certain components, or have a different arrangement of components.
In one embodiment, there is also provided a storage medium storing a computer program comprising program instructions which, when executed by a computer, cause the computer to perform the method according to the preceding embodiment, the computer may be part of the mining and recommending apparatus for potential investment enterprises mentioned above.
It will be understood by those skilled in the art that all or part of the processes of the methods of the embodiments described above can be implemented by a computer program, which can be stored in a computer-readable storage medium, and when executed, can include the processes of the embodiments of the methods described above. The storage medium may be a magnetic disk, an optical disk, a Read-Only Memory (ROM), a Random Access Memory (RAM), or the like.
It will be apparent to those skilled in the art that the modules or steps of the invention described above may be implemented in a general purpose computing device, they may be centralized on a single computing device or distributed across a network of computing devices, and optionally they may be implemented in program code executable by a computing device, such that they may be stored on a computer storage medium (ROM/RAM, magnetic disks, optical disks) and executed by a computing device, and in some cases, the steps shown or described may be performed in an order different than that described herein, or they may be separately fabricated into individual integrated circuit modules, or multiple ones of them may be fabricated into a single integrated circuit module. Thus, the present invention is not limited to any specific combination of hardware and software.
The foregoing is a more detailed description of the present invention that is presented in conjunction with specific embodiments, and the practice of the invention is not to be considered limited to those descriptions. For those skilled in the art to which the invention pertains, several simple deductions or substitutions can be made without departing from the spirit of the invention, and all shall be considered as belonging to the protection scope of the invention.

Claims (7)

1. A method for mining and recommending potential investment enterprises is characterized by comprising the following steps:
designing an investment intention label common to enterprises according to the constituent factors of the investment intention label, and setting a potential label value formula;
establishing an initial submerged excavation model based on a limit gradient lifting algorithm according to the investment wish label;
obtaining historical success data, and training the initial submerged entry mining model to obtain a target submerged entry mining model;
acquiring characteristics of a target enterprise, wherein the characteristics are used for representing basic data of the target enterprise;
predicting the investment willingness of the target enterprise according to the characteristics and the target latent investment mining model to obtain a predicted target willingness value;
and recommending the target enterprises with the target willingness values higher than the preset threshold value.
2. The method of claim 1, wherein the potential tag value formula is specifically:
label＝wDEALDEAL+wINCORRINCORR+wSORELASORELA+wCIINTERCIINTER+
wVENDINCVENDINC+wARGARG+wPROGROPROGRO+wRARGRORATGRO+
wASSGROASSGRO+wINNOVAINNOVA+wINDEXPINDEXP+wCOMEXPCOMEXP。
3. the method of claim 2, wherein the label is willingness to invest label, DEAL indicates whether or not successful parking has been performed, and is a discrete value, and successful parking is recorded as 1, otherwise, it is recorded as 0; INCORR represents enterprise investment relevance, which is also a discrete value, and the relation is marked as 1, otherwise, the relation is marked as 0; the SORELA represents the social relevance degree, is a discrete value and has a relevance of 1, otherwise, the social relevance degree is 0; CIINTER represents the enterprise industry association degree, and has an association of 1, otherwise, 0;
VENDINC, ARG, PROGRO, ratgo, assro, inova, INDEXP, and COMEXP respectively represent normalized scores and value ranges (0, 1) of an enterprise in terms of "main revenue", "annual ring ratio growth", "profit sum", "tax sum", "asset sum", "innovation capacity", "industry expandability", and "enterprise expandability", respectively;
wherein wi(i ═ VENDING, ARG, PROGRO, RATGRO, ASSGRO, INNOVA, INDEXP, COMEXP) respectively represent the linear weights of the above indexes in order, and satisfy Σ wi＝1。
4. The method of claim 3, wherein said label satisfies 0 ≦ label ≦ 1.
5. The utility model provides a mining and recommendation device of potential investment enterprise, which comprises a label design module, a model building module, a model training module, a characteristic obtaining module, a will forecasting module and an enterprise recommendation module, wherein:
the label design module is used for designing an investment intention label universal for an enterprise according to the constituent factors of the investment intention label and setting a potential label value formula;
the model establishing module is used for establishing an initial submerged excavation model based on a limit gradient lifting algorithm according to the investment willingness label;
the model training module is used for acquiring historical success data and training the initial submerged excavation model to obtain a target submerged excavation model;
the characteristic acquisition module is used for acquiring the characteristics of the target enterprise, and the characteristics are used for representing basic data of the target enterprise;
the intention prediction module is used for predicting the investment intention of the target enterprise according to the characteristics and the target latent investment mining model to obtain a predicted target intention value;
the enterprise recommending module is used for recommending the target enterprises with the target willingness values higher than the preset threshold value.
6. An apparatus comprising a memory, a processor and a computer program stored on the memory and executable on the processor, characterized in that the steps of the method of any of claims 1 to 4 are implemented when the computer program is executed by the processor.
7. A storage medium having a computer program stored thereon, the computer program, when being executed by a processor, realizing the steps of the method of any one of claims 1 to 4.
CN202011177090.3A 2020-10-29 2020-10-29 Mining and recommending method, device, equipment and storage medium for potential investment enterprises Pending CN112381653A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202011177090.3A CN112381653A (en) 2020-10-29 2020-10-29 Mining and recommending method, device, equipment and storage medium for potential investment enterprises

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202011177090.3A CN112381653A (en) 2020-10-29 2020-10-29 Mining and recommending method, device, equipment and storage medium for potential investment enterprises

Publications (1)

Publication Number Publication Date
CN112381653A true CN112381653A (en) 2021-02-19

Family

ID=74577855

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202011177090.3A Pending CN112381653A (en) 2020-10-29 2020-10-29 Mining and recommending method, device, equipment and storage medium for potential investment enterprises

Country Status (1)

Country Link
CN (1) CN112381653A (en)

Cited By (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN113342904A (en) * 2021-04-01 2021-09-03 山东省人工智能研究院 Enterprise service recommendation method based on enterprise feature propagation
CN116091241A (en) * 2023-02-10 2023-05-09 广州市城市规划勘测设计研究院 Investment prediction method based on complex different composition
CN118350691A (en) * 2024-03-28 2024-07-16 北京清华同得科技有限公司 Evaluation method and system for regional intelligent introduction enterprises

Cited By (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN113342904A (en) * 2021-04-01 2021-09-03 山东省人工智能研究院 Enterprise service recommendation method based on enterprise feature propagation
CN116091241A (en) * 2023-02-10 2023-05-09 广州市城市规划勘测设计研究院 Investment prediction method based on complex different composition
CN118350691A (en) * 2024-03-28 2024-07-16 北京清华同得科技有限公司 Evaluation method and system for regional intelligent introduction enterprises

Similar Documents

Publication Publication Date Title
CN112381653A (en) Mining and recommending method, device, equipment and storage medium for potential investment enterprises
CN110310012B (en) Data analysis method, device, equipment and computer readable storage medium
Cao et al. Early warning of enterprise decline in a life cycle using neural networks and rough set theory
Zhong et al. Effects of cost-benefit analysis under back propagation neural network on financial benefit evaluation of investment projects
García de Soto et al. A process for the development and evaluation of preliminary construction material quantity estimation models using backward elimination regression and neural networks
Wu et al. [Retracted] Optimization and Simulation of Enterprise Management Resource Scheduling Based on the Radial Basis Function (RBF) Neural Network
Wang et al. RETRACTED ARTICLE: Investor behaviour monitoring based on deep learning
US20220036387A1 (en) Multi-faceted large-scale forecasting
Yurynets et al. Neural Network Modeling of the Social and Economic, Investment and Innovation Policy of the State.
Ilich et al. Modeling of employment structural transformations
Co et al. Comparison between ARIMA and LSTM-RNN for VN-index prediction
Siraj et al. Pre-selection of recruitment candidates using case based reasoning
Xiao et al. A dynamic multi-model transfer based short-term load forecasting
Shostak et al. Computer support for decision-making on defining the strategy of Green IT development at the state level
Pant et al. Forecasting of nifty 50 index volatility using seasonal ARIMA model
Smerichevskyi et al. Managing the potential of digital transformation of Ukraine’s economy
CN114742657A (en) Investment target planning method and system
Lemke Probabilistic energy forecasting based on self-organizing inductive modeling
TW202121157A (en) Method and quantifying a data based on final value and estimate
TWI419071B (en) Active knowledge management system, method and computer program product for problem solving
Xu et al. A time series combined forecasting model based on prophet-lgbm
Martelo et al. Multivariate Financial Time Series Forecasting with Deep Learning
Yao Towards a Better Forecasting Model for Economic Indices.
US20230146635A1 (en) Method and Systems for Conditioning Data Sets for Efficient Computational Processing
Pierrot et al. On tracking varying bounds when forecasting bounded time series

Legal Events

Date Code Title Description
PB01 Publication
PB01 Publication
SE01 Entry into force of request for substantive examination
SE01 Entry into force of request for substantive examination
WD01 Invention patent application deemed withdrawn after publication
WD01 Invention patent application deemed withdrawn after publication

Application publication date: 20210219