CN110097216A - For the prediction technique and device of enterprise development, server - Google Patents
For the prediction technique and device of enterprise development, server Download PDFInfo
- Publication number
- CN110097216A CN110097216A CN201910290504.4A CN201910290504A CN110097216A CN 110097216 A CN110097216 A CN 110097216A CN 201910290504 A CN201910290504 A CN 201910290504A CN 110097216 A CN110097216 A CN 110097216A
- Authority
- CN
- China
- Prior art keywords
- prediction
- enterprise
- data
- learning model
- machine learning
- 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
Links
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
- 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
- G06Q10/063—Operations research, analysis or management
-
- 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
- G06Q10/063—Operations research, analysis or management
- G06Q10/0639—Performance analysis of employees; Performance analysis of enterprise or organisation operations
-
- 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
Landscapes
- Business, Economics & Management (AREA)
- Engineering & Computer Science (AREA)
- Human Resources & Organizations (AREA)
- Strategic Management (AREA)
- Economics (AREA)
- Entrepreneurship & Innovation (AREA)
- Development Economics (AREA)
- Marketing (AREA)
- Game Theory and Decision Science (AREA)
- Operations Research (AREA)
- Quality & Reliability (AREA)
- Tourism & Hospitality (AREA)
- Physics & Mathematics (AREA)
- General Business, Economics & Management (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Educational Administration (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
This application discloses a kind of prediction technique for enterprise development and devices, server.This method includes obtaining default machine learning model according to the training of known business data;Unknown input Target Enterprise data are to the default machine learning model;By the default machine learning model, the prediction result for the unknown object enterprise is exported.Present application addresses the analyzing and diagnosings of enterprise development to need the technical issues of relying on mass data.It is realized by the application and is realized based on machine learning model to the enterprise development progress predictive diagnosis in industry, unknown input Target Enterprise data are found into similar enterprises by default machine learning model, and the Gernral Check-up and development prediction result to unknown object enterprise are generated according to similar enterprises track.In addition, application itself can also carry out prediction result to keep track optimization, iteration updates default machine learning model.
Description
Technical field
This application involves enterprise diagnosis, personnel management field, in particular to a kind of prediction side for enterprise development
Method and device, server.
Background technique
Forecast analysis is carried out to enterprise development, it usually needs by expert or advisory organization, analyze mass data
It diagnoses and obtain enterprise development predicts result.
Inventors have found that needing to expend a large amount of manpower and financial resources, and the standard predicted to the analyzing and diagnosing of enterprise's mass data
True rate is not high.Further, it cannot achieve and keep track.
Aiming at the problem that analyzing and diagnosing of enterprise development in the related technology needs to rely on mass data, not yet propose have at present
The solution of effect.
Summary of the invention
The main purpose of the application is to provide a kind of prediction technique for enterprise development and device, server, with
Solve the problems, such as that the analyzing and diagnosing of enterprise development needs to rely on mass data.
To achieve the goals above, according to the one aspect of the application, a kind of prediction side for enterprise development is provided
Method.
The prediction technique for enterprise development according to the application includes: to obtain default machine according to the training of known business data
Device learning model;Unknown input Target Enterprise data are to the default machine learning model;Pass through the default machine learning mould
Type exports the prediction result for the unknown object enterprise.
Further, by the default machine learning model, the development prediction for the unknown object enterprise is exported
As a result, the development prediction result includes at least: industry development scale forecast, employee development scale forecast.
Further, by the default machine learning model, the diagnosis prediction for the unknown object enterprise is exported
As a result, the diagnosis prediction result includes at least: similar period diagnosis prediction result, similar size diagnosis prediction result, similar
Talent's diagnosis prediction result, similar financial diagnosis prediction result, analogous tissue's structural diagnosis prediction result.
Further, the default machine learning model uses full Connection Neural Network model.
Further, obtaining default machine learning model according to the training of known business data includes:
Known business data are determined in target industry;
According to the known business data, filters out unknown object and examine enterprise and known business data according to preset rules
Data combination.
Further, by the default machine learning model, the prediction result for the unknown object enterprise is exported
Later further include:
Judge whether the prediction result meets default prediction tendency;
If it is determined that the prediction result does not meet default prediction tendency, then tracking is adjusted described by intervention operation
Prediction result, and continue to carry out data backflow.
To achieve the goals above, according to the another aspect of the application, a kind of prediction dress for enterprise development is provided
It sets.
It include: training module according to the prediction meanss for enterprise development of the application, for according to known business data
Training obtains default machine learning model;Input module, for unknown input Target Enterprise data to the default machine learning
Model;Prediction module, for exporting the prediction knot for the unknown object enterprise by the default machine learning model
Fruit.
Further, the training module comprises determining that unit, for determining known business data in target industry;
Screening unit, for filtering out unknown object and examining enterprise and known business data according to default according to the known business data
The data combination of rule.
Further, the prediction module includes: judging unit, for judging it is default pre- whether the prediction result meets
Survey tendency;Intervene unit, for when judging that the prediction result does not meet default prediction tendency, then tracking to pass through intervention operation
The prediction result adjusted, and continue to carry out data backflow.
To achieve the goals above, according to the another aspect of the application, providing a kind of server includes: the prediction
Device.
It is used for the prediction technique and device of enterprise development in the embodiment of the present application, using according to the training of known business data
The mode for obtaining default machine learning model is reached by unknown input Target Enterprise data to the default machine learning model
It has arrived through the default machine learning model, has exported the purpose of the prediction result for the unknown object enterprise, thus real
Show the technical effect realized based on machine learning model and carry out predictive diagnosis to the enterprise development in industry, and then solves enterprise
The analyzing and diagnosing of industry development needs the technical issues of relying on mass data.
Detailed description of the invention
The attached drawing constituted part of this application is used to provide further understanding of the present application, so that the application's is other
Feature, objects and advantages become more apparent upon.The illustrative examples attached drawing and its explanation of the application is for explaining the application, not
Constitute the improper restriction to the application.In the accompanying drawings:
Fig. 1 is according to the prediction technique flow diagram for enterprise development in the application first embodiment;
Fig. 2 is according to the prediction technique flow diagram for enterprise development in the application second embodiment;
Fig. 3 is according to the prediction technique flow diagram for enterprise development in the application 3rd embodiment;
Fig. 4 is according to the prediction meanss structural schematic diagram for enterprise development in the application first embodiment;
Fig. 5 is according to the prediction meanss structural schematic diagram for enterprise development in the application second embodiment;
Fig. 6 is according to the prediction meanss structural schematic diagram for enterprise development in the application 3rd embodiment.
Fig. 7 is principle configuration diagram.
Specific embodiment
In order to make those skilled in the art more fully understand application scheme, below in conjunction in the embodiment of the present application
Attached drawing, the technical scheme in the embodiment of the application is clearly and completely described, it is clear that described embodiment is only
The embodiment of the application a part, instead of all the embodiments.Based on the embodiment in the application, ordinary skill people
Member's every other embodiment obtained without making creative work, all should belong to the model of the application protection
It encloses.
It should be noted that the description and claims of this application and term " first " in above-mentioned attached drawing, "
Two " etc. be to be used to distinguish similar objects, without being used to describe a particular order or precedence order.It should be understood that using in this way
Data be interchangeable under appropriate circumstances, so as to embodiments herein described herein.In addition, term " includes " and " tool
Have " and their any deformation, it is intended that cover it is non-exclusive include, for example, containing a series of steps or units
Process, method, system, product or equipment those of are not necessarily limited to be clearly listed step or unit, but may include without clear
Other step or units listing to Chu or intrinsic for these process, methods, product or equipment.
It should be noted that in the absence of conflict, the features in the embodiments and the embodiments of the present application can phase
Mutually combination.The application is described in detail below with reference to the accompanying drawings and in conjunction with the embodiments.
The prediction technique for enterprise development of the embodiment of the present application, including preset according to the training of known business data
Machine learning model;Unknown input Target Enterprise data are to the default machine learning model;Pass through the default machine learning
Model exports the prediction result for the unknown object enterprise.Unknown input target is looked forward to by default machine learning model
Industry data find similar enterprises, and generate the Gernral Check-up and development prediction to unknown object enterprise according to similar enterprises track
As a result.
As shown in Figure 1, this method includes the following steps, namely S102 to step S102:
Step S102 obtains default machine learning model according to the training of known business data;
Data extract source and include known business public information data and preset full link data in known business data.Its
In, the data volume of the known business public information data presets full link data greater than described.But preset full link data
Data depth be greater than the known business public information data.It is described preset full link data establish exclusive data and
Intervene or non-fiddle factors weight, data are more accurate.
It is described that preset full link data include but is not limited to scope of the enterprise data, the affiliated industry data of enterprise, business finance
Data, enterprise personnel data, organization structure of the enterprise data etc..
It should be noted that the not source to specific known business data or data structure in embodiments herein
It is defined, as long as can satisfy training obtains the requirement of default machine learning model, those skilled in the art can root
Factually border service condition is selected or is configured.
It is also to be noted that the machine learning model is based on full Connection Neural Network model, the engineering
It practises model training module and multiple learning training is carried out to known industry and enterprise data, obtain the prediction processing result of Target Enterprise.
Specific full Connection Neural Network model structure is not defined in embodiments herein, as long as can satisfy study
Model can find the similar enterprise similar with Target Enterprise.
Step S104, unknown input Target Enterprise data to the default machine learning model;
Unknown object business data may include: unknown industry unknown object business data or known industry unknown object enterprise
Industry data.By carrying out multiple learning training to the default machine learning model using known industry and enterprise data, accomplish mesh
Enterprise's Gernral Check-up and development prediction are marked, and solution is provided and keeps track optimization.It additionally can use unknown industry enterprise
Industry data are trained the similar enterprise of identification according to the similarity for presetting dimension (for example, staff size etc.) in full link data is preset
Industry is good for.
In addition, the machine learning model training module passes through the similarity meter of default dimension to unknown industry and enterprise data
It calculates, obtains the prediction processing result of similar enterprise.Precision is identified to unknown corporate model so as to improve.
Step S106 exports the prediction result for the unknown object enterprise by the default machine learning model.
The Gernral Check-up provided by the default machine learning model according to similar period, similar size, the similar talent,
Similar financial, analogous tissue's structure, next year development trend etc..
After providing solution by Gernral Check-up or prediction development, sent out using the enterprise of manual intervention under line and non-intervention
Tendency comparison diagnosis is opened up, judges whether the development trend for meeting known business sample data, does not need to intervene if meeting, such as
Fruit does not meet, and is intervened.
Preferably, the solution provided according to Gernral Check-up is provided by the default machine learning model, carries out line
Lower manual intervention or the Developing Trend track data not carried out under line after personnel's intervention keep track optimization data backflow.It realizes
The enterprise development tracking optimization of manual intervention after solution is provided to diagnosis.
It can be seen from the above description that the application realizes following technical effect:
It is used for the prediction technique of enterprise development in the embodiment of the present application, it is pre- using being obtained according to the training of known business data
If the mode of machine learning model, by unknown input Target Enterprise data to the default machine learning model, reach logical
The default machine learning model is crossed, the purpose of the prediction result for the unknown object enterprise is exported, to realize base
The technical effect for carrying out predictive diagnosis to the enterprise development in industry is realized in machine learning model, and then solves enterprise development
Analyzing and diagnosing need the technical issues of relying on mass data.
As preferred in the present embodiment, by the default machine learning model, the unknown object is looked forward in output
The development prediction of industry is as a result, the development prediction result includes at least: industry development scale forecast, employee development scale forecast.
The development prediction for the unknown object enterprise is exported as a result, simultaneously by artificial intelligence according to the default machine learning model
And included at least in the development prediction result, unknown enterprises ' industry development scale is predicted.Unknown enterprise personnel is developed
Scale forecast.The prediction of industry development rule is the important indicator of forecasting enterprise development, and Qua-ntile Regression prediction is also enterprise's hair
Open up the important indicator of prediction.Simultaneously by the prediction result to industry development and employee development, it is also used as whether enterprise is good for
The diagnosis important indicator of health.
As preferred in the present embodiment, by the default machine learning model, the unknown object is looked forward in output
The diagnosis prediction of industry is as a result, the diagnosis prediction result includes at least: similar period diagnosis prediction result, similar size diagnosis are pre-
Survey result, similar talent's diagnosis prediction result, similar financial diagnosis prediction result, analogous tissue's structural diagnosis prediction result.It is logical
Cross similar period diagnosis prediction result, similar size diagnosis prediction result, similar talent's diagnosis prediction result, similar financial diagnosis
Prediction result, analogous tissue's structural diagnosis prediction result can carry out comprehensive diagnostic to the health of enterprise, while pass through similar enterprise
The similar period diagnosis prediction result of industry, similar size diagnosis prediction result, similar talent's diagnosis prediction result, similar financial are examined
Disconnected prediction result, analogous tissue's structural diagnosis prediction result, can also predict the development prediction situation of Target Enterprise.
As preferred in the present embodiment, the default machine learning model uses full Connection Neural Network model.It needs
It is noted that the full Connection Neural Network model in the embodiment of the present application can use existing model structure, in the reality of the application
It applies in example and without specifically limiting, as long as can satisfy the requirement of machine learning.
According to the embodiment of the present application, as preferred in the present embodiment, as shown in Fig. 2, according to the training of known business data
Obtaining default machine learning model includes:
Step S202 determines known business data in target industry;
Step S204, according to the known business data, filter out unknown object examine enterprise and known business data according to
The data of preset rules combine.
Specifically, for needing to judge whether known sample enterprise belongs to single row as machine learning model input data
Industry or multiple industries, if Target Enterprise belongs to single industry, data rule of combination, which may is that, will need to diagnose Target Enterprise
It establishes and matches in such a way that scale, finance, talent's label, place city, scale of investment, market are occupied etc. with comparison business data respectively
To syntagmatic, learning model is input to after carrying out pairing rules combination.If it is determined that known business sample belongs to multiple industries
Then relevant industries are carried out with the combination of said combination regular fashion.
According to the embodiment of the present application, as preferred in the present embodiment, as shown in figure 3, passing through the default machine learning
Model, after output is for the prediction result of the unknown object enterprise further include:
Step S302, judges whether the prediction result meets default prediction tendency;
Step S304, if it is determined that the prediction result does not meet default prediction tendency, then tracking passes through intervention operation tune
The prediction result after whole, and continue to carry out data backflow.
Specifically, develop in solution offer module by dual model comparison diagnosis, in the Gernral Check-up and prediction
Experts database recommends Adjusted Option, in the improvement for keeping track and returning in optimization module and comparing improvement and be worth sustainable iteration.
According to the embodiment of the present application, as preferred in the present embodiment, for enterprise development prediction technique, comprising:
Default machine learning model is obtained according to the training of known business data;
Unknown input Target Enterprise data are to the default machine learning model;
By the default machine learning model, the prediction result for the unknown object enterprise is exported.
By the default machine learning model, the development prediction for the unknown object enterprise is exported as a result, described
Development prediction result includes at least: industry development scale forecast, employee development scale forecast.
The diagnosis prediction for the unknown object enterprise is exported as a result, the diagnosis prediction result includes at least: similar
Period diagnosis prediction result, similar size diagnosis prediction result, similar talent's diagnosis prediction result, similar financial diagnosis prediction knot
Fruit, analogous tissue's structural diagnosis prediction result.
The default machine learning model uses full Connection Neural Network model.
Obtaining default machine learning model according to the training of known business data includes:
Known business data are determined in target industry;
According to the known business data, filters out unknown object and examine enterprise and known business data according to preset rules
Data combination.
By the default machine learning model, exports and the prediction result of the unknown object enterprise is also wrapped later
It includes:
Judge whether the prediction result meets default prediction tendency;
If it is determined that the prediction result does not meet default prediction tendency, then tracking is adjusted described by intervention operation
Prediction result, and continue to carry out data backflow.
It should be noted that step shown in the flowchart of the accompanying drawings can be in such as a group of computer-executable instructions
It is executed in computer system, although also, logical order is shown in flow charts, and it in some cases, can be with not
The sequence being same as herein executes shown or described step.
According to the embodiment of the present application, additionally provide a kind of for implementing the prediction dress for enterprise development of the above method
It sets, as shown in figure 4, the device includes: training module 10, for obtaining default machine learning mould according to the training of known business data
Type;Input module 20, for unknown input Target Enterprise data to the default machine learning model;Prediction module 30, is used for
By the default machine learning model, the prediction result for the unknown object enterprise is exported.
It includes that known business discloses that data, which extract source, in known business data in the training module 10 of the embodiment of the present application
Information data and preset full link data.Wherein, the data volume of the known business public information data is greater than described default complete
Link data.But the data depth for presetting full link data is greater than the known business public information data.Described default
It is more accurate that full link data establishes exclusive data and intervention or non-fiddle factors weight, data.
It is described that preset full link data include but is not limited to scope of the enterprise data, the affiliated industry data of enterprise, business finance
Data, enterprise personnel data, organization structure of the enterprise data etc..
It should be noted that the not source to specific known business data or data structure in embodiments herein
It is defined, as long as can satisfy training obtains the requirement of default machine learning model, those skilled in the art can root
Factually border service condition is selected or is configured.
It is also to be noted that the machine learning model is based on full Connection Neural Network model, the machine learning mould
Type training module carries out multiple learning training to known industry and enterprise data, obtains the prediction processing result of Target Enterprise.At this
Specific full Connection Neural Network model structure is not defined in the embodiment of application, as long as can satisfy learning model
The similar enterprise similar with Target Enterprise can be found.
Unknown object business data may include: unknown industry unknown object enterprise in the input module 20 of the embodiment of the present application
Industry data or known industry unknown object business data.By utilizing known industry and enterprise data to the default machine learning mould
Type carries out multiple learning training, accomplishes Target Enterprise Gernral Check-up and development prediction, and provide solution and keep track optimization.
It additionally can use unknown industry and enterprise data and preset dimension (for example, staff size etc.) according to presetting in full link data
Similarity is trained the similar enterprise of identification and is good for.
In addition, the machine learning model training module passes through the similarity meter of default dimension to unknown industry and enterprise data
It calculates, obtains the prediction processing result of similar enterprise.Precision is identified to unknown corporate model so as to improve.
The Gernral Check-up foundation provided in the prediction module 30 of the embodiment of the present application by the default machine learning model
Similar period, similar size, the similar talent, similar financial, analogous tissue's structure, next year development trend etc..
After providing solution by Gernral Check-up or prediction development, sent out using the enterprise of manual intervention under line and non-intervention
Tendency comparison diagnosis is opened up, judges whether the development trend for meeting known business sample data, does not need to intervene if meeting, such as
Fruit does not meet, and is intervened.
Preferably, the solution provided according to Gernral Check-up is provided by the default machine learning model, carries out line
Lower manual intervention or the Developing Trend track data not carried out under line after personnel's intervention keep track optimization data backflow.It realizes
The enterprise development tracking optimization of manual intervention after solution is provided to diagnosis.
According to the embodiment of the present application, as preferred in the present embodiment, as shown in figure 5, the training module comprises determining that
Unit 101, for determining known business data in target industry;Screening unit 102, for according to the known business number
According to, filter out unknown object examine enterprise and known business data according to preset rules data combine.
Specifically, for needing to judge whether known sample enterprise belongs to single row as machine learning model input data
Industry or multiple industries, if Target Enterprise belongs to single industry, data rule of combination, which may is that, will need to diagnose Target Enterprise
It establishes and matches in such a way that scale, finance, talent's label, place city, scale of investment, market are occupied etc. with comparison business data respectively
To syntagmatic, learning model is input to after carrying out pairing rules combination.If it is determined that known business sample belongs to multiple industries
Then relevant industries are carried out with the combination of said combination regular fashion.
According to the embodiment of the present application, as preferred in the present embodiment, as shown in fig. 6, the prediction module 30 includes: to sentence
Disconnected unit 301, for judging whether the prediction result meets default prediction tendency;Intervene unit 302, for described in the judgement
When prediction result does not meet default prediction tendency, then tracking is by the intervention operation prediction result adjusted, and continue into
Row data backflow.
Specifically, develop in solution offer module by dual model comparison diagnosis, in the Gernral Check-up and prediction
Experts database recommends Adjusted Option, in the improvement for keeping track and returning in optimization module and comparing improvement and be worth sustainable iteration.
According to another embodiment of the application, a kind of server is additionally provided, comprising: the prediction meanss for enterprise development.
The realization principle and beneficial effect of the prediction meanss are for example above-mentioned, are no longer repeated herein.
As shown in fig. 7, the realization principle of the application is as follows:
Firstly, leaning on the empirical tradition side of " expert " and " consulting firm " entirely by changing original system of business management adjustment
Method characterizes the health condition of enterprise using health indicator.It specifically, include different data dimension and specific association
Action data.Then, artificial intelligence learning model is recycled to be recommended and returned.Enterprise's desensitization number of magnanimity based on tens of thousands of
Different scales, the business data parameter normal range (NR) of different industries are provided according to (not being related to the data of privacy of user).In the enterprise
In industry data parameter ranges, data modeling, dual model comparison diagnosis are carried out to specific enterprise repeatedly.And experts database under lead-in
Recommend to be adjusted scheme, return to compare again and improve optimizing for the sustainable iteration being worth.
In addition, by learning model have apparent network effects and synergistic effect, if be added SaaS enterprise
Node is more, scale is more neat, and effect data is better.So carrying out data model iteration, Ke Yiwei by distinguishing different industries
Relevant industries enterprise is more focused more effective solution.
By the method in the embodiment of the present application, enterprise can be made from problem definition, request for proposal and recruitment evaluation,
There can be data can be according to there is method that can use for reference.To which adjustment promotes internal competition power quickly, to promote externally service power.
Specifically include following module:
Business data extraction module, collaborative filtering calculate scoring modules, machine learning model training module, Gernral Check-up and
Prediction develops solution and provides module, keeps track optimization module.Be specifically related to: business data extracts, machine learning model
Training, Gernral Check-up and development prediction, provide solution and keep track optimization data backflow.
Wherein, by repeatedly to above-mentioned specific enterprise carry out in the machine learning model training module data modeling,
Develop solution in the Gernral Check-up and prediction and provides dual model comparison diagnosis in module, in the Gernral Check-up and prediction
Development solution provides experts database in module and recommends Adjusted Option, keeps track recurrence comparison improvement in optimization module described
It is worth the improvement of sustainable iteration.
Targeted diagnostics enterprise during the business data extraction module, the learning training of machine learning model and right
It is extracted from known business data sample library than enterprise sample, and relevant industries can be chosen every time and included at least but unlimited
Model learning training is carried out in 5 enterprise's samples.For needing to judge that known sample is looked forward to as machine learning model input data
Whether industry belongs to single industry or multiple industries, if Target Enterprise belongs to single industry, data rule of combination may is that by
Need to diagnose Target Enterprise respectively with comparison business data with scale, finance, talent's label, place city, scale of investment, market
The modes such as occupy and establish combinations of pairs relationship, is input to learning model after carrying out pairing rules combination.If it is determined that known business
Sample belongs to the combination that multiple industries then carry out said combination regular fashion to relevant industries.
In the business data extraction module, data extract source and include known business public information data and preset full chain
Circuit-switched data.Wherein, the data volume of the known business public information data presets full link data greater than described.But it is default complete
The data depth of link data is greater than the known business public information data.It is described preset full link data establish it is exclusive
Data and intervention or non-fiddle factors weight, data are more accurate.
It include but is not limited to scope of the enterprise data, the affiliated industry data of enterprise, enterprise in the business data extraction module
Industry financial data, enterprise personnel data, organization structure of the enterprise data etc..
The similarity marking that the collaborative filtering calculates scoring modules is based on collaborative filtering.
The machine learning model training module carries out multiple learning training to known industry and enterprise data, obtains target enterprise
The prediction processing result of industry.
The machine learning model training module by the similarity calculation of default dimension, obtains unknown industry and enterprise data
To the prediction processing result of similar enterprise.Precision is identified to unknown corporate model so as to improve.Pass through the engineering
Model training is practised, Gernral Check-up and development prediction result are exported.By repeatedly inputting Target Enterprise data to machine learning model
After can export affinity score and search similar business data, analyzed according to similar enterprise diagnosis, prediction develops the offer solution
Scheme, and keep track optimization data backflow finally provides the Developing Trend data that manual intervention whether is carried out after solution
Preset full link data is flowed back into again carries out Continuous optimization.
The Gernral Check-up and prediction develop solution provide module, the Gernral Check-up that can be provided according to it is similar when
Phase, similar size, the similar talent, similar financial, analogous tissue's structure, next year development trend etc..Pass through Gernral Check-up
Or after prediction development provides solution, compared and diagnosed using the enterprise development tendency of manual intervention under line and non-intervention, judgement
The development trend for whether meeting known business sample data does not need to intervene if meeting, be intervened if not meeting.
It is described to keep track optimization module by the solution that provides according to Gernral Check-up, carry out under line manual intervention or
The Developing Trend track data after personnel intervene under line is not carried out keeps track optimization data backflow.It realizes and provides solution to diagnosis
Certainly optimization is tracked in the enterprise development of manual intervention after scheme.
Machine learning model training module, firstly, examining enterprise and comparison using target is chosen in existing industry and enterprise data
The combination of enterprise's preset data rule, the data as machine learning model input, and carry out engineering to the machine learning model
Practise training.Then, it identifies that the similar enterprises of the Target Enterprise carry out identifying and diagnosing by machine learning model, and passes through class
Development track like enterprise predicts the Target Enterprise health status and future thrust and provides instruction.
Further, the machine learning model is based on full Connection Neural Network model, and in known business data sample
Increase the full link data of enterprise (including but not limited to, scope of the enterprise, equity structure, founder, institutional framework, the talent distribution,
Attendance, operation cost etc.).
Further, selected targeted diagnostics enterprise sample includes at least five in relevant industries difference business data, is passed through
Several business data are input in machine learning model before diagnosis Target Enterprise industry contrast sample calculates similarity highest automatically
It is ranked up according to comprehensive similarity score;Target Enterprise is subjected to different business data rule groups by score value height sequence later
It closes, thus retains multiple sample rules combination that the comparison business data sample inputs original similarity score ordering rule,
These regular combined sample data are constituted into a matrix fraction by sequencing of similarity again, as the input of machine learning model,
Last machine learning model exports corresponding marking.Learnt by training, the score of the machine learning model output can be made
Correspond to the relevant enterprise of industry.And then it can use known industry and enterprise data and repeatedly study instruction carried out to machine learning model
Practice, accomplishes Target Enterprise Gernral Check-up and development prediction, and solution is provided and keeps track optimization.It additionally can use not
Knowing and doing industry business data is trained knowledge according to the similarity for presetting dimension (for example, staff size etc.) in full link data is preset
Not similar enterprise carries out Gernral Check-up and development prediction, finally provides solution, keeps track optimization and effect returns.
By method provided by the present application, and by actual measurement, in the medium-sized and small enterprises sample experiment of 173 part of 200~500 people
In, it is more than 98.3% in the effective percentage that three Ge Yuenei enterprises are obviously improved, significantly enhances Enterprise Experts ability, hence it is evident that reduces enterprise
The decision-making period of industry.
Obviously, those skilled in the art should be understood that each module of above-mentioned the application or each step can be with general
Computing device realize that they can be concentrated on a single computing device, or be distributed in multiple computing devices and formed
Network on, optionally, they can be realized with the program code that computing device can perform, it is thus possible to which they are stored
Be performed by computing device in the storage device, perhaps they are fabricated to each integrated circuit modules or by they
In multiple modules or step be fabricated to single integrated circuit module to realize.In this way, the application be not limited to it is any specific
Hardware and software combines.
The foregoing is merely preferred embodiment of the present application, are not intended to limit this application, for the skill of this field
For art personnel, various changes and changes are possible in this application.Within the spirit and principles of this application, made any to repair
Change, equivalent replacement, improvement etc., should be included within the scope of protection of this application.
Claims (10)
1. a kind of prediction technique for enterprise development characterized by comprising
Default machine learning model is obtained according to the training of known business data;
Unknown input Target Enterprise data are to the default machine learning model;
By the default machine learning model, the prediction result for the unknown object enterprise is exported.
2. prediction technique according to claim 1, which is characterized in that pass through the default machine learning model, output pair
In the unknown object enterprise development prediction as a result, the development prediction result includes at least: industry development scale forecast, people
Member's development scale prediction.
3. prediction technique according to claim 1, which is characterized in that pass through the default machine learning model, output pair
In the unknown object enterprise diagnosis prediction as a result, the diagnosis prediction result includes at least: similar period diagnosis prediction knot
Fruit, similar size diagnosis prediction result, similar talent's diagnosis prediction result, similar financial diagnosis prediction result, analogous tissue's knot
Structure diagnosis prediction result.
4. prediction technique according to claim 1, which is characterized in that the default machine learning model is using full connection mind
Through network model.
5. prediction technique according to claim 1, which is characterized in that obtain default machine according to the training of known business data
Learning model includes:
Known business data are determined in target industry;
According to the known business data, filters out unknown object and examine enterprise and known business data according to the data of preset rules
Combination.
6. prediction technique according to claim 1, which is characterized in that pass through the default machine learning model, output pair
After the prediction result of the unknown object enterprise further include:
Judge whether the prediction result meets default prediction tendency;
If it is determined that the prediction result does not meet default prediction tendency, then tracking passes through the intervention operation prediction adjusted
As a result, and continuing to carry out data backflow.
7. a kind of prediction meanss for enterprise development characterized by comprising
Training module, for obtaining default machine learning model according to the training of known business data;
Input module, for unknown input Target Enterprise data to the default machine learning model;
Prediction module, for exporting the prediction result for the unknown object enterprise by the default machine learning model.
8. prediction meanss according to claim 7, which is characterized in that the training module includes:
Determination unit, for determining known business data in target industry;
Screening unit, for according to the known business data, filter out unknown object examine enterprise and known business data according to
The data of preset rules combine.
9. prediction meanss according to claim 7, which is characterized in that the prediction module includes:
Judging unit, for judging whether the prediction result meets default prediction tendency;
Intervene unit, for when judging that the prediction result does not meet default prediction tendency, then tracking to pass through intervention operation tune
The prediction result after whole, and continue to carry out data backflow.
10. a kind of server characterized by comprising prediction meanss as claimed in claim 7 or 8.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201910290504.4A CN110097216A (en) | 2019-04-11 | 2019-04-11 | For the prediction technique and device of enterprise development, server |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201910290504.4A CN110097216A (en) | 2019-04-11 | 2019-04-11 | For the prediction technique and device of enterprise development, server |
Publications (1)
Publication Number | Publication Date |
---|---|
CN110097216A true CN110097216A (en) | 2019-08-06 |
Family
ID=67444708
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201910290504.4A Pending CN110097216A (en) | 2019-04-11 | 2019-04-11 | For the prediction technique and device of enterprise development, server |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN110097216A (en) |
Cited By (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN111080132A (en) * | 2019-12-18 | 2020-04-28 | 北京智识企业管理咨询有限公司 | Industry chain analysis system and method based on big data |
CN111222774A (en) * | 2019-12-30 | 2020-06-02 | 广州博士信息技术研究院有限公司 | Enterprise data analysis method and device and server |
CN113269516A (en) * | 2021-05-13 | 2021-08-17 | 企家有道网络技术(北京)有限公司 | Method, device and system for improving enterprise energy efficiency through big data |
CN113269515A (en) * | 2021-05-13 | 2021-08-17 | 企家有道网络技术(北京)有限公司 | Enterprise service method, device and system based on three-order six-step system |
CN116542801A (en) * | 2023-07-07 | 2023-08-04 | 成都理工大学 | Financial data analysis method and system |
Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN107193994A (en) * | 2017-06-07 | 2017-09-22 | 前海梧桐(深圳)数据有限公司 | Business decision point method for digging and its system based on mass data |
CN107392456A (en) * | 2017-07-14 | 2017-11-24 | 武汉理工大学 | A kind of multi-angle rating business credit modeling method for merging internet information |
CN107844901A (en) * | 2017-11-01 | 2018-03-27 | 前海梧桐(深圳)数据有限公司 | A kind of method and its system of enterprise operation automated analysis |
CN109102140A (en) * | 2018-06-11 | 2018-12-28 | 北交金科金融信息服务有限公司 | A kind of growing quality evaluation method of base artificial intelligence and big data technology |
CN109345089A (en) * | 2018-09-13 | 2019-02-15 | 杭州索骥数据科技有限公司 | Enterprise development state evaluating method and system based on big data |
-
2019
- 2019-04-11 CN CN201910290504.4A patent/CN110097216A/en active Pending
Patent Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN107193994A (en) * | 2017-06-07 | 2017-09-22 | 前海梧桐(深圳)数据有限公司 | Business decision point method for digging and its system based on mass data |
CN107392456A (en) * | 2017-07-14 | 2017-11-24 | 武汉理工大学 | A kind of multi-angle rating business credit modeling method for merging internet information |
CN107844901A (en) * | 2017-11-01 | 2018-03-27 | 前海梧桐(深圳)数据有限公司 | A kind of method and its system of enterprise operation automated analysis |
CN109102140A (en) * | 2018-06-11 | 2018-12-28 | 北交金科金融信息服务有限公司 | A kind of growing quality evaluation method of base artificial intelligence and big data technology |
CN109345089A (en) * | 2018-09-13 | 2019-02-15 | 杭州索骥数据科技有限公司 | Enterprise development state evaluating method and system based on big data |
Cited By (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN111080132A (en) * | 2019-12-18 | 2020-04-28 | 北京智识企业管理咨询有限公司 | Industry chain analysis system and method based on big data |
CN111222774A (en) * | 2019-12-30 | 2020-06-02 | 广州博士信息技术研究院有限公司 | Enterprise data analysis method and device and server |
CN113269516A (en) * | 2021-05-13 | 2021-08-17 | 企家有道网络技术(北京)有限公司 | Method, device and system for improving enterprise energy efficiency through big data |
CN113269515A (en) * | 2021-05-13 | 2021-08-17 | 企家有道网络技术(北京)有限公司 | Enterprise service method, device and system based on three-order six-step system |
CN116542801A (en) * | 2023-07-07 | 2023-08-04 | 成都理工大学 | Financial data analysis method and system |
CN116542801B (en) * | 2023-07-07 | 2023-09-22 | 成都理工大学 | Financial data analysis method and system |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN110097216A (en) | For the prediction technique and device of enterprise development, server | |
Nieto et al. | Supporting academic decision making at higher educational institutions using machine learning-based algorithms | |
CN110084493A (en) | Enterprise diagnosis, prediction technique and device, server based on artificial intelligence | |
Gonçalves et al. | Multi-criteria decision methodology for selecting maintenance key performance indicators | |
Zhang et al. | A quantitative approach to design alternative evaluation based on data-driven performance prediction | |
Morard et al. | Time evolution analysis and forecast of key performance indicators in a balanced scorecard | |
CN110110898A (en) | Based on the industry analysis method and device of enterprise's health indicator, server | |
Minku et al. | How to make best use of cross-company data for web effort estimation? | |
Tian et al. | A fuzzy TOPSIS model via chi-square test for information source selection | |
CN110335168A (en) | Method and system based on GRU optimization power information acquisition terminal fault prediction model | |
CN114548494A (en) | Visual cost data prediction intelligent analysis system | |
Ailenei | Process mining tools: A comparative analysis | |
Yu et al. | Multi-criteria decision making for PLM maturity analysis based on an integrated fuzzy AHP and VIKOR methodology | |
Ziegenbein et al. | Machine learning algorithms in machining: A guideline for efficient algorithm selection | |
CN115660296B (en) | Automatic issuing method of non-compliance project proposal scheme based on machine learning | |
Paracchini et al. | Weighting and aggregation of indicators for sustainability impact assessment in the SENSOR context | |
CN110084411A (en) | Learning model and implementation method for predicting to return | |
Silva et al. | IT project investments: an analysis based on a sort and rank problem | |
Rostek | Modeling commercial potential of innovative projects | |
CN114493226A (en) | Efficiency evaluation system for company operation management based on multidimensional angle analysis | |
Tsehayae et al. | Data-driven approaches to discovering knowledge gaps related to factors affecting construction labor productivity | |
Parkavi et al. | Predicting the course knowledge level of students using data mining techniques | |
Sari et al. | E-Supply Chain Management Readiness Analysis by Using the Smart Industry Readiness Index (Case Study: Pt Kuliner Dapur Bersama) | |
Liu et al. | Data-driven based student programming competition award prediction via machine learning models | |
Harman et al. | Simple rules outperform Machine Learning in the 3rd annual SIOP ML competition. |
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 | ||
RJ01 | Rejection of invention patent application after publication |
Application publication date: 20190806 |
|
RJ01 | Rejection of invention patent application after publication |