CN106326473A - Data mining method based on entropy weight algorithm and analytic hierarchy process and system thereof - Google Patents
Data mining method based on entropy weight algorithm and analytic hierarchy process and system thereof Download PDFInfo
- Publication number
- CN106326473A CN106326473A CN201610791303.9A CN201610791303A CN106326473A CN 106326473 A CN106326473 A CN 106326473A CN 201610791303 A CN201610791303 A CN 201610791303A CN 106326473 A CN106326473 A CN 106326473A
- Authority
- CN
- China
- Prior art keywords
- evaluation index
- index
- evaluation
- enterprise
- matrix
- 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
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/20—Information retrieval; Database structures therefor; File system structures therefor of structured data, e.g. relational data
- G06F16/24—Querying
- G06F16/245—Query processing
- G06F16/2458—Special types of queries, e.g. statistical queries, fuzzy queries or distributed queries
- G06F16/2465—Query processing support for facilitating data mining operations in structured databases
-
- 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
- G06Q10/06393—Score-carding, benchmarking or key performance indicator [KPI] analysis
Landscapes
- Engineering & Computer Science (AREA)
- Business, Economics & Management (AREA)
- Human Resources & Organizations (AREA)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Databases & Information Systems (AREA)
- Strategic Management (AREA)
- Entrepreneurship & Innovation (AREA)
- Economics (AREA)
- Educational Administration (AREA)
- Development Economics (AREA)
- General Physics & Mathematics (AREA)
- Software Systems (AREA)
- Fuzzy Systems (AREA)
- Data Mining & Analysis (AREA)
- Computational Linguistics (AREA)
- Probability & Statistics with Applications (AREA)
- Mathematical Physics (AREA)
- Game Theory and Decision Science (AREA)
- General Engineering & Computer Science (AREA)
- Marketing (AREA)
- Operations Research (AREA)
- Quality & Reliability (AREA)
- Tourism & Hospitality (AREA)
- General Business, Economics & Management (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention provides a data mining method based on entropy weight algorithm and analytic hierarchy process, which is characterized by including the following steps: S1, collecting original data used for enterprise evaluations; S2, establishing enterprise evaluation systems from four aspects: technology levels of enterprises, competitive capacities of enterprises, economic values of enterprises and social values of enterprises, and determining evaluation indexes; S3, using entropy weight algorithm and analytic hierarchy process to determine index weights of each evaluation index in enterprise evaluations; S4, according to the evaluation indexes and the corresponding index weights, using fuzzy comprehensive evaluation algorithm to conduct enterprise evaluations and get the results of the enterprise evaluations. By establishing an all-around enterprise evaluation system, the data-mining method can evaluate the technical advancements and practical values of enterprises. And by combining entropy weight algorithm, analytic hierarchy process and fuzzy comprehensive evaluation algorithm, the data-mining method makes the results of the enterprise evaluations more objective and scientific, thus revealing more accurately the strengths of science and technology capabilities of enterprises and offering important references for enterprises to make periodic strategic objectives.
Description
Technical field
The present invention relates to technical field of data processing, be specifically related to a kind of data based on entropy weight algorithm Yu analytic hierarchy process (AHP)
Method for digging and system.
Background technology
Enterprise generally requires during development and carries out substantial amounts of business activity, in order to improves the popularity of enterprise, promotees
Enter the further development of enterprise.
But, as cross enterprise to self conditions of the enterprise can not clearly, understand accurately if, enterprise is in development
During be likely to due to mistake information and cause incorrect decision, have a strong impact on the development of enterprise.
Owing to needing enterprise is carried out omnibearing analysis, investigation during carrying out valuation of enterprise, and wherein can relate to
And the process to substantial amounts of data message, what this was the biggest improves complexity and the inconvenience of valuation of enterprise.
Meanwhile, in the mass data information related to during being directed to valuation of enterprise, obtain for enterprise's different aspect
The different data message taken, during valuation of enterprise, it often has different importances, and i.e. some data message can
Important proportion can be occupied, and some data message may be the most secondary, the most effective for different pieces of information
The importance of information has the valuation of enterprise that carries out of proportion, the most further improves the difficulty of valuation of enterprise.
Therefore, how to provide a kind of to realize enterprise is carried out full side for the company information with different weight
The valuation of enterprise method of the evaluation of position just becomes problem demanding prompt solution.
Summary of the invention
The invention provides a kind of data digging method based on entropy weight algorithm and analytic hierarchy process (AHP) and system, by setting up
Valuation of enterprise system, determines the weight of business evaluation indicator so that valuation of enterprise is more in conjunction with entropy weight algorithm and analytic hierarchy process (AHP)
Add comprehensive and accurate.
The Part I of the present invention provides a kind of data digging method based on entropy weight algorithm Yu analytic hierarchy process (AHP), including
Following steps:
S1, gather for the initial data of valuation of enterprise；
S2, with science and technology in enterprise level, enterprise competitiveness, business economic is worth and corporate social is worth four aspects and sets up
Valuation of enterprise system, and determine evaluation index；
S3, employing entropy weight algorithm and analytic hierarchy process (AHP) determine the index weights of each evaluation index in valuation of enterprise；
S4, fuzzy overall evaluation algorithm is used to carry out valuation of enterprise according to the index weights of standard diagrams and correspondence thereof,
To valuation of enterprise result.
Preferably, in S3, entropy weight algorithm comprises the following steps:
S31, evaluation index in initial data is carried out matrix normalization process；
S32, the evaluation index processed through matrix normalization is carried out entropy definition；
S33, entropy according to evaluation index determine the entropy weight of evaluation index；
The middle-level analytic process of S3 comprises the following steps:
S34, layer rank relation according to evaluation index, set up evaluation index Recurison order hierarchy system；
S35, according to evaluation index Recurison order hierarchy system is positioned at the evaluation index of same stratum, set up evaluation index two-by-two
Multilevel iudge matrix, and the relative importance of evaluation index is determined according to judgment matrix；
S36, for the low stratum evaluation index corresponding to ibid stratum's evaluation index, relatively heavy according to evaluation index
The property wanted, carries out Mode of Level Simple Sequence to evaluation index, determines the same order weight of evaluation index；
S37, according to the same order weight of low stratum evaluation index and the same order weight of upper stratum's evaluation index of correspondence thereof,
Evaluation index is carried out total hierarchial sorting, determines the layer rank weight of evaluation index；
Entropy weight according to evaluation index and layer rank weight, calculate described index weights by linear weighted function averaging method.
It is further preferred that the matrix normalization process of evaluation index calculates by the following method in S31:
S311, setting total m evaluation object, n evaluation index, then raw data matrix is A=(aij)m×n, wherein, A
For raw data matrix, aijFor i-th evaluation object, the initial data that jth evaluation index is corresponding；
S312, raw data matrix is carried out matrix normalization process after, through matrix normalization process after original number
It is R=(r according to matrixij)m×n, wherein, R is the raw data matrix after matrix normalization processes, rijFor i-th evaluation object,
The normalization data that jth evaluation index is corresponding；
S313, when evaluation index is forward index, evaluation index normalization formula is:
When evaluation index is negative sense index, evaluation index normalization formula is:
Define by with lower section it is further preferred that the evaluation index processed through matrix normalization is carried out entropy by S32
Method calculates:
If total m evaluation object, n evaluation index, wherein the formula of the entropy of jth evaluation index is:
Wherein, hjFor i-th evaluation object, the entropy of jth evaluation index, k is the standardized data of m evaluation object,
fijFor i-th evaluation object, the normalization data ratio that jth evaluation index is corresponding；
And,
It is further preferred that the entropy weight of evaluation index calculates by the following method in S33:
If total m evaluation object, n evaluation index, wherein the formula of the entropy weight of jth evaluation index is:
Wherein, wjEntropy weight for jth evaluation index.
It is further preferred that in S36 before carrying out Mode of Level Simple Sequence, the concordance of test and judge matrix；
When judging that judgment matrix meets concordance, then evaluation index is carried out Mode of Level Simple Sequence；
When judging that judgment matrix is unsatisfactory for concordance, then adjust the judgment matrix of correspondence.
It is further preferred that according to Mode of Level Simple Sequence, S36 determines that same order weight calculates by the following method:
For judgment matrix B, calculate BW=λmaxThe characteristic root of W and characteristic vector；
To the characteristic vector normalization process calculated, obtain the eigenvectors matrix W=[w after normalization1, w2...
wn,]r, and using the characteristic vector after this normalization as the same order weight of evaluation index；
Wherein, W is the eigenvectors matrix after normalization, λmaxIt is characterized root, wnBe feature after the n-th normalization to
Amount, r is characterized moment of a vector rank of matrix.
It is further preferred that linear weighted function averaging method is:
Index weights=entropy weight × weight × 0.5,0.5+ layer rank.
Preferably, carry out valuation of enterprise according to the index weights of standard diagrams and correspondence thereof in 43 to comprise the following steps:
S41, according to valuation of enterprise Establishing business evaluation indicator collection and Comment gathers, and determine according to expert opinion data
With reference to Evaluations matrix；
S42, according to index set and with reference to Evaluations matrix use fuzzy overall evaluation algorithm evaluation index is evaluated, obtain
Obtain fuzzy evaluating matrix；
S43, draw fuzzy evaluation results according to the index weights of evaluation index and fuzzy evaluating matrix, and comment according to fuzzy
Valency result comment of correspondence in Comment gathers determines valuation of enterprise result.
The Part II of the present invention provides a kind of data digging system based on entropy weight algorithm Yu analytic hierarchy process (AHP), bag
Include:
Achievement data acquisition terminal, it gathers the initial data for valuation of enterprise；
Appraisement system builds server, and it is worth with science and technology in enterprise level, enterprise competitiveness, business economic for building
It is worth four aspects with corporate social and sets up valuation of enterprise system, and determine evaluation index；
Data processing server, it builds server with achievement data acquisition terminal and appraisement system respectively and is connected, according to
Evaluation index is processed by initial data by entropy weight algorithm and analytic hierarchy process (AHP), in order to evaluation index index weigh
Weight；
Valuation of enterprise server, it is connected with data processing server, comments according to the reference of evaluation index set in advance
Valence mumber evidence, in conjunction with corresponding index weights, calculates valuation of enterprise data.
Preferably, data processing server includes:
Entropy weight algorithm process unit, it is for by the entropy weight of entropy weight algorithm Calculation Estimation index；
Analytic hierarchy process (AHP) processing unit, it is for by the layer rank weight of analytic hierarchy process (AHP) Calculation Estimation index；
Index weights computing unit, it is for the entropy weight according to index weights and layer rank weight calculation index weights.
It is further preferred that entropy weight algorithm process unit includes:
Normalized subelement, it is for carrying out matrix normalization process to evaluation index in initial data；
Entropy definition subelement, it is for carrying out entropy definition to the evaluation index through normalized；
Entropy weight computation subunit, it for determining the entropy weight of evaluation index according to the entropy of evaluation index.
It is further preferred that normalized subelement, evaluation index is normalized by by the following method:
If total m evaluation object, n evaluation index, then raw data matrix is A=(aij)m×n, wherein, A is original
Data matrix, aijFor i-th evaluation object, the initial data that jth evaluation index is corresponding；
Raw data matrix after raw data matrix is carried out matrix normalization process, after matrix normalization processes
For R=(rij)m×n, wherein, R is the raw data matrix after matrix normalization processes, rijFor i-th evaluation object, jth is commented
The normalization data that valency index is corresponding；
When evaluation index is forward index, evaluation index normalization formula is:
When evaluation index is negative sense index, evaluation index normalization formula is:
It is further preferred that entropy definition subelement, it carries out entropy definition by the following method to evaluation index:
If total m evaluation object, n evaluation index, wherein the formula of the entropy of jth evaluation index is:
Wherein, hjFor i-th evaluation object, the entropy of jth evaluation index, k is the standardized data of m evaluation object,
fijFor i-th evaluation object, the normalization data ratio that jth evaluation index is corresponding；
And,
It is further preferred that entropy weight computation subunit, the entropy weight of its Calculation Estimation index by the following method:
If total m evaluation object, n evaluation index, wherein the formula of the entropy weight of jth evaluation index is:
Wherein, wjEntropy weight for jth evaluation index.
It is further preferred that analytic hierarchy process (AHP) processing unit includes:
Evaluation index Recurison order hierarchy system construction subelement, it is set up evaluate for the layer rank relation according to evaluation index
Index Recurison order hierarchy system；
Judgment matrix builds subelement, and it is for according to the evaluation being positioned at same stratum in evaluation index Recurison order hierarchy system
Index, sets up evaluation index multilevel iudge matrix two-by-two, and determines the relative importance of evaluation index according to judgment matrix；
Consistency detection subelement, it is for the concordance of test and judge matrix；
Same order weight calculation subelement, it refers to for evaluating for the low stratum corresponding to ibid stratum's evaluation index
Mark, according to the relative importance of evaluation index, carries out Mode of Level Simple Sequence to evaluation index, determines the same order weight of evaluation index；
Layer rank weight calculation subelement, it is for the same order weight according to low stratum evaluation index and the upper single order of correspondence thereof
The same order weight of layer evaluation index, carries out total hierarchial sorting to evaluation index, determines the layer rank weight of evaluation index.
It is further preferred that same order weight calculation subelement, the same order weight of its Calculation Estimation index by the following method:
For judgment matrix B, calculate BW=λmaxThe characteristic root of W and characteristic vector；
To the characteristic vector normalization process calculated, obtain the eigenvectors matrix W=[w after normalization1, w2...
wn,]r, and using the characteristic vector after normalization as the same order weight of evaluation index；
Wherein, W is the eigenvectors matrix after normalization, λmaxIt is characterized root, wnBe feature after the n-th normalization to
Amount, r is characterized moment of a vector rank of matrix.
It is further preferred that index weights computing unit, the index weights of its Calculation Estimation index by the following method:
Index weights=entropy weight × weight × 0.5,0.5+ layer rank.
Preferably, valuation of enterprise server includes:
Appraisement system construction unit, it is used for building valuation of enterprise system, and the index set of correspondence and Comment gathers；
Fuzzy algorithmic approach processing unit, its for by fuzzy overall evaluation algorithm according to the index weights of evaluation index and ginseng
Examine evaluating data and calculate valuation of enterprise data；
Evaluation result unit, it is for according to Comment gathers, it is thus achieved that the valuation of enterprise result that valuation of enterprise data are corresponding；
Display unit, it is used for showing output valuation of enterprise result.
The data digging method based on entropy weight algorithm and analytic hierarchy process (AHP) of the present invention and system thereof, comprehensive by setting up
Valuation of enterprise system, technical advance and practical value to enterprise are evaluated, and are divided with level by entropy weight algorithm simultaneously
Analysis method and the combination of Field Using Fuzzy Comprehensive Assessment so that valuation of enterprise result is more objective, science, in order to represent more accurately
Go out the power of science and technology in enterprise ability, formulate interim strategic objective for enterprise and important references is provided.
Accompanying drawing explanation
In order to be illustrated more clearly that the technical scheme of the embodiment of the present invention, required use in embodiment being described below
Accompanying drawing be briefly described, it should be apparent that, below describe in accompanying drawing be some embodiments of the present invention, for ability
From the point of view of the those of ordinary skill of territory, on the premise of not paying creative work, it is also possible to obtain other according to these accompanying drawings
Accompanying drawing.
Fig. 1 is present invention flow process based on entropy weight algorithm Yu an embodiment of the data digging method of analytic hierarchy process (AHP)
Figure.
Fig. 2 is that present invention entropy weight based on entropy weight algorithm with an embodiment of the data digging method of analytic hierarchy process (AHP) is calculated
Method flow chart.
Fig. 3 is that present invention level based on entropy weight algorithm with an embodiment of the data digging method of analytic hierarchy process (AHP) divides
Analysis method flow chart.
Fig. 4 is that present invention enterprise based on entropy weight algorithm with an embodiment of the data digging method of analytic hierarchy process (AHP) is commented
Valency flow chart.
Fig. 5 is present invention structure based on entropy weight algorithm Yu an embodiment of the data digging system of analytic hierarchy process (AHP)
Figure.
Detailed description of the invention
For making the object, technical solutions and advantages of the present invention clearer, attached below in conjunction with in the embodiment of the present invention
Figure, is clearly and completely described the technical scheme in the embodiment of the present invention, it is clear that described embodiment is the present invention
A part of embodiment rather than whole embodiments.Based on the embodiment in the present invention, those of ordinary skill in the art are not having
Make the every other embodiment obtained under creative work premise, broadly fall into the scope of protection of the invention.
Embodiment one
Fig. 1 is present invention flow process based on entropy weight algorithm Yu an embodiment of the data digging method of analytic hierarchy process (AHP)
Figure；Fig. 2 is present invention entropy weight algorithm stream based on entropy weight algorithm Yu an embodiment of the data digging method of analytic hierarchy process (AHP)
Cheng Tu；Fig. 3 is present invention step analysis based on entropy weight algorithm Yu an embodiment of the data digging method of analytic hierarchy process (AHP)
Method flow chart；Fig. 4 is present invention enterprise based on entropy weight algorithm Yu an embodiment of the data digging method of analytic hierarchy process (AHP)
Evaluation rubric figure.
As it is shown in figure 1, in the present embodiment, data digging method based on entropy weight algorithm Yu analytic hierarchy process (AHP), including following
Step:
S1, gather for the initial data of valuation of enterprise.
Concrete, for the required enterprise carrying out valuation of enterprise, carry out the collection of the initial data for valuation of enterprise, with
Continue the carrying out of valuation of enterprise after an action of the bowels.
S2, with science and technology in enterprise level, enterprise competitiveness, business economic is worth and corporate social is worth four aspects and sets up
Valuation of enterprise system, and determine evaluation index.
Concrete, with science and technology in enterprise level, enterprise competitiveness, business economic is worth and corporate social is worth four evaluations
The initial data of enterprise is classified by classification, will the initial data of enterprise carry out respectively according to four different classifications
Analyzing and processing.
It is worth and corporate social value four in science and technology in enterprise level, enterprise competitiveness, business economic for evaluating enterprise
Evaluate the performance of classification, set up valuation of enterprise system for these four classifications, meanwhile, determine and valuation of enterprise system relates to
Evaluation index.
Wherein, the evaluation index related in valuation of enterprise system can further include three grades of indexs, two-level index and
First class index, with this by evaluation index is divided into three stratum, it is achieved the evaluation index of three stratum from small to large, from point
Global scope assessment enterprise to face.
Further, first class index may particularly include: science and technology in enterprise level, enterprise competitiveness, business economic be worth and
Corporate social is worth, i.e. four evaluation aspects of valuation of enterprise system.
Two-level index may particularly include: Corporation R & D ability, Intellectual Property Right of Enterprises, senior enterprise leader information, enterprise believe substantially
Breath, corporate financial information, enterprises information, enterprise investment information, entreprise cost information and enterprise's case information.
Three grades of indexs may particularly include: R & D Cost ratio, research staff's ratio, enterprise patent quantity, enterprise patent divide
Class quantity, enterprise patent percent of pass, software copyright, enterprise founders educational background level, enterprise founders foundation experience, enterprise's wound
Beginning people's working experience level, enterprise's shareholder's entirety educational background, enterprise set up time, enterprise employees number, enterprise's turnover, section of enterprise
Grind personnel ratios, enterprise marketing total value, enterprise assets total value, corporate profit margin, property right of enterprises ratio, business economic value added,
Break one's promise information, Enterprise Business Management State, enterprise branch office's information, enterprise of enterprises registration capital, enterprise owes loan information, enterprise investment
The amount of money, business entity's investments abroad amount of money, asset-liability ratio, running cost rate, Company Knowledge are externally speculated by mechanism's quantity, enterprise
Property right case information, enterprise's criminal case information, enterprise's civil case information, enterprise's administrative case information and enterprise perform case
Information.
Further, two-level index is the further refinement for first class index, and three grades of indexs are to enter one for two-level index
Step refinement, and in evaluation index, the three mutual corresponding relations of level are the most as shown in table 1.
Table 1
S3, employing entropy weight algorithm and analytic hierarchy process (AHP) determine the index weights of each evaluation index in valuation of enterprise.
Concrete, for the evaluation index determined in S2, come according in S1 step by entropy weight algorithm and analytic hierarchy process (AHP)
The initial data gathered calculates the index weights of each evaluation index, in order to the follow-up weight information according to evaluation index comes enterprise
It is evaluated.
As in figure 2 it is shown, the present embodiment uses the entropy weight of entropy weight algorithm Calculation Estimation index may particularly include following steps:
S31, evaluation index in initial data is carried out matrix normalization process.
Concrete, the initial data that the evaluation index in untreated enterprise's initial data is relevant is carried out matrix and returns
One change processes.
Further, the matrix normalization of evaluation index processes and can calculate by the following method:
S311, set total m evaluation object, n evaluation index, then initial data in the valuation of enterprise system of the present embodiment
Matrix is as shown in Equation 1:
Formula 1
A=(aij)m×n；
Wherein, A is raw data matrix, aijFor i-th evaluation object, the initial data that jth evaluation index is corresponding.
S312, raw data matrix is normalized after, through matrix normalization process after initial data square
Battle array as shown in Equation 2:
Formula 2
R=(rij)m×n；
Wherein, R is the raw data matrix after matrix normalization processes, rijFor i-th evaluation object, jth evaluation refers to
The normalization data that mark is corresponding.
S313, when the evaluation index carrying out matrix normalization process is forward index, i.e. this evaluation index numerical value is greatly
Time excellent, then the normalization formula of this evaluation index is as shown in Equation 3:
Formula 3
And when the evaluation index carrying out matrix normalization process is negative sense index, i.e. this evaluation index numerical value is little for excellent
Time, evaluation index normalization formula as shown in Equation 4:
Formula 4
S32, the evaluation index processed through matrix normalization is carried out entropy definition.
Further, the evaluation index processed through matrix normalization carries out entropy definition calculate by the following method:
According to m evaluation object in valuation of enterprise system, n evaluation index, the wherein entropy of jth evaluation index such as formula
Shown in 5:
Formula 5
Wherein, hjFor the entropy of jth evaluation index, k is the standardized data of m evaluation object, fijIt is right to evaluate for i-th
As, the normalization data ratio that jth evaluation index is corresponding；
Further, in formula 5, the value of k and fij can be respectively:
S33, entropy according to evaluation index determine the entropy weight of evaluation index；
Further, the entropy weight of evaluation index can calculate by the following method:
For the situation in the present embodiment, i.e. have m evaluation object when valuation of enterprise system, during n evaluation index, its
Shown in the formula equation below 6 of the entropy weight of middle jth evaluation index:
Formula 6
Wherein, wjEntropy weight for jth evaluation index.
The present embodiment use the layer rank weight flow process of analytic hierarchy process (AHP) Calculation Estimation index as it is shown on figure 3, it specifically wraps
Include following steps:
S34, layer rank relation according to evaluation index, set up evaluation index Recurison order hierarchy system.
Further, according to the layer rank relation of evaluation index involved in the valuation of enterprise system set up in S2, set up
The index Recurison order hierarchy system of evaluation index, i.e. first class index in the present embodiment of described above, two-level index and three grades refer to
Respective layer rank relation between mark.
S35, according to evaluation index Recurison order hierarchy system is positioned at the evaluation index of same stratum, set up evaluation index two-by-two
Multilevel iudge matrix, and the relative importance of evaluation index is determined according to judgment matrix.
It is positioned at the evaluation index in same stratum in index Recurison order hierarchy system, contrasts two-by-two, and set up phase
The judgment matrix of two corresponding evaluation indexes, such as, carries out the foundation of judgment matrix, or two two grades between two three grades of indexs
Carry out carrying out between the foundation of judgment matrix, or two first class index the foundation of judgment matrix between index.
According to the above explanation to the present embodiment, when there being m evaluation object, during n evaluation index, for importance
Evaluation result AkFor, evaluation index BiAnd BjThe judgment matrix B of relative importanceijAs shown in table 2:
Table 2
| Ak | B1 | B2 | Λ | Bn |
| B1 | B11 | B12 | Λ | B1n |
| B2 | B21 | B22 | Λ | B2n |
| M | M | M | M | |
| Bn | Bn1 | Bn2 | Λ | Bnm |
Being directed to above judgment matrix, it should meet following relation:
Formula 7
Bij=1, i, j=1,2,3, n and i=j；
Formula 8
I, j=1,2,3, n and i ≠ j；
Wherein, BijFor evaluation index BiAnd BjJudgment matrix, bijBe i-th to evaluation index relative to jth to evaluation index
Judgment matrix scale.
As shown in table 2, BijRepresent the significance level of the evaluation index that the evaluation index of the i-th row arranges, and i relative to jth
Be 1,2,3, n, j be 1,2,3, m.
And, in table 2 relatively important sexual relationship generally take 1,2,3,9 and their inverse as judgment matrix
Scale represents, in order to quickly obtain the relative importance between two evaluatings, this reality by judgment matrix scale
Execute the implication of judgment matrix scale and correspondence thereof in example the most as shown in table 3:
Table 3
The relative importance between evaluation index is finally determined by above method.
Further, for the judgement of relative importance between evaluation index two-by-two, can access method, expert consult by inquiry
Inquiry method is evaluated the determination of index, and according to the importance Judgement Matricies of each evaluation index, carries out calculating phase with this
To importance.
S36, for the low stratum evaluation index corresponding to ibid stratum's evaluation index, relatively heavy according to evaluation index
The property wanted, carries out Mode of Level Simple Sequence to evaluation index, determines the same order weight of evaluation index.
Further, before evaluation index is carried out Mode of Level Simple Sequence, in addition it is also necessary to the concordance of test and judge matrix, with really
Determine whether evaluation index can be carried out Mode of Level Simple Sequence, ensure the follow-up accuracy to valuation of enterprise result with this.
Judgment matrix carries out consistency check to be judged by equation below 9:
Formula 9
Wherein, CI is judgment matrix approach index, λmaxBeing characterized root, n is the number of evaluation index in judgment matrix.
When calculating CI=0 according to formula 9, then judge that this judgment matrix has concordance, and this judgment matrix has
Crash consistency, follow-up can carry out Mode of Level Simple Sequence to the evaluation index that this judgment matrix is corresponding.
And according to formula 9, as λ in formula 9maxThe value of-n is the biggest, then the value of CI is the biggest, then is judged as this and sentences
Disconnected matrix does not have crash consistency, and the concordance of this judgment matrix is poor.
And for the poor situation of the concordance of judgment matrix, the most also need to carry out CI with Aver-age Random Consistency Index RI
Relatively, carry out the concordance judging judgment matrix the most definitely with this, and whether can be straight according to this judgment matrix
Tap into row Mode of Level Simple Sequence.
Wherein, in the present embodiment, the value of Aver-age Random Consistency Index RI is as shown in table 4:
Table 4
| Exponent number n | 1 | 2 | 3 | 4 | 5 | 6 | 7 | 8 | 9 |
| RI | 0.00 | 0.00 | 0.58 | 0.90 | 1.12 | 1.24 | 1.32 | 1.41 | 1.45 |
Judge whether judgment matrix has concordance by the relation of CI and RI, specifically can be judged by formula 10:
Formula 10
CR=CI/RI < 0.10；
Wherein, CR is the coincident indicator criterion of judgment matrix.
As CR meets the relation of formula 10, then judge that this judgment matrix has satisfied concordance；And if CR is unsatisfactory for
The relation of formula 10, then judge that this judgment matrix does not possess concordance and it needs to readjust judgment matrix, and to warp
Cross the judgment matrix after adjusting again to judge whether to meet concordance.
And for Mode of Level Simple Sequence, it can calculate especially by the following manner:
For judgment matrix B, calculate characteristic root and the characteristic vector of equation 1 below 1:
Formula 11
BW=λmaxW；
Wherein, W is the eigenvectors matrix after normalization, λmaxIt is characterized root.
Further, carry out normalization process for the characteristic vector calculated, obtain the such as formula of the characteristic vector after normalization
Shown in 12:
Formula 12
W=[w1, w2... wn,]r；
Wherein, wnBeing the characteristic vector after the n-th normalization, r is characterized moment of a vector rank of matrix.
Finally, using the characteristic vector after this normalization as the same order weight of evaluation index.
Concrete, for the above method judging same order weight according to judgment matrix, for four one of evaluation enterprise
It is as follows that level index carries out building Evaluations matrix process:
If in first class index, science and technology in enterprise level compare science and technology in enterprise level, enterprise competitiveness, business economic be worth and
The judgment matrix scale that corporate social is worth is respectively 1,2,3,4, then build Evaluations matrix based on this as shown in table 5:
Table 5
Further, in upper table 5, including four first class index, i.e. n=4 in the present embodiment altogether, in n th Root, Section 1 can root
Calculate according to below equation 21:
Formula 21
Can obtain Section 1 in n th Root according to formula 21 is 0.452, and can calculate in an Evaluations matrix according to the method
The value of each in n th Root row, i.e. relative importance.
And, according to the relative importance drawn, finally determine four first class index by same order weight.
S37, according to the same order weight of low stratum evaluation index and the same order weight of upper stratum's evaluation index of correspondence thereof,
Evaluation index is carried out total hierarchial sorting, determines the layer rank weight of evaluation index.
Concrete, according to the same order weight of evaluation index in the low stratum calculated, corresponding to this evaluation index
The same order weight of evaluation index in a upper stratum, carries out total hierarchial sorting with secondary to all of evaluation index, determines with this
Go out each evaluation index with reference to the layer rank weight after stratum's evaluation index on it.
Further, before carrying out layer calculating of rank weight, also need evaluation index correspondence same order weights at different levels composition
Matrix carries out consistency check.
Finally, determined the entropy weight of evaluation index by entropy weight algorithm, and also determine evaluation by analytic hierarchy process (AHP)
The layer rank weight of index, can carry out the index weights of Calculation Estimation index according to this entropy weight and layer rank weight, and this index weights can be adopted
Calculate with linear weighted function averaging method.
Further, in the present embodiment, for using calculation linear weighted function to put down by the entropy weight of evaluation index and layer rank weight
The method that all algorithm carrys out parameter weight, can use equation below 13 to calculate:
Formula 13
Index weights=entropy weight × weight × 0.5,0.5+ layer rank.
S4, fuzzy overall evaluation algorithm is used to carry out valuation of enterprise according to the index weights of standard diagrams and correspondence thereof,
To valuation of enterprise result.
According to the index weights of the evaluation index calculated, and the evaluation index corresponding to this index weights, by fuzzy
Comprehensive evaluation algorithm evaluates enterprise, and finally draws valuation of enterprise result.
Wherein, as shown in Figure 4, carry out valuation of enterprise according to the index weights of standard diagrams and correspondence thereof and may also include following
Step:
S41, according to valuation of enterprise Establishing business evaluation indicator collection and Comment gathers, and determine according to expert opinion data
With reference to Evaluations matrix.
For the present embodiment, if the index set of valuation of enterprise system is O, and O can be represented by formula 14:
Formula 14
O={P1,P2,P3,P4}；
Wherein, P1To P4For each first class index.
Further, P1, P2, P3, P4 are represented by formula 15:
Formula 15
P1={ Q1,Q2,Q3}
P2={ Q4,Q5}
P3={ Q6,Q7}
P1={ Q8,Q9}
Wherein, Q1To Q9For each two-level index.
And, the Comment gathers of the present embodiment for for enterprise technology characterization and evaluation be qualified, domestic general, domestic medium,
Domestic advanced person, domestically leading totally 5 grades, and this Comment gathers can be indicated by formula 16:
Formula 16
V={v1,v2,v3,v4,v5,={ 1,2,3,4,5}；
Wherein, v1、v2、v3、v4And v5Respectively corresponding advanced person medium, domestic general, domestic qualified, domestic, domestically leading five
Individual grade.
For with reference to Evaluations matrix, the present embodiment specifically have chosen 50 experts according to a certain science and technology in enterprise technical merit
Evaluating data, obtain following with reference to Evaluations matrix:
Wherein, R1’、R2’、R3' and R4' it being respectively the reference Evaluations matrix for first class index expert, a, b, c and d are respectively
Reference Evaluations matrix value scale for the two-level index expert corresponding for first class index.
S42, according to index set and with reference to Evaluations matrix use fuzzy overall evaluation algorithm evaluation index is evaluated, obtain
Obtain fuzzy evaluating matrix.
In the present embodiment, two-level index in evaluation index is carried out fuzzy overall evaluation algorithm, is then commented by fuzzy synthesis
The fuzzy evaluating matrix that valency algorithm calculates is as shown in Equation 17:
Formula 17
Wherein, R1For fuzzy evaluating matrix, it draws based on evaluation index and with reference to Evaluations matrix, r11、r12、r13、r14、
r15For the scale in fuzzy evaluating matrix.
Further, r in formula 1711、r12、r13、r14、r15Value as shown in formula 18:
Formula 18
r11={ Q1*a11+Q2*a21+Q3*a31}；r12={ Q1*a12+Q2*a22+Q3*a32}.
r13={ Q1*a13+Q2*a23+Q3*a33}；r14={ Q1*a14+Q2*a24+Q3*a34}.
r15={ Q1*a15+Q2*a25+Q3*a35}.；
In like manner, R can also be obtained by above formula 17 and formula 182、R3And R4Fuzzy evaluating matrix, until will be all
Two-level index is after all fuzzy overall evaluation algorithm has calculated fuzzy evaluating matrix, then the fuzzy evaluating matrix of first order calculation index.
S43, draw fuzzy evaluation results according to the index weights of evaluation index and fuzzy evaluating matrix, and comment according to fuzzy
Valency result comment of correspondence in Comment gathers determines valuation of enterprise result.
Calculate fuzzy evaluation results by index weights and fuzzy evaluating matrix to be calculated by equation below 19:
Formula 19
M=W × R={m1,m2,m3,m4}；
Wherein, M is fuzzy evaluation results, and W is index weights, and R is fuzzy evaluating matrix, m1、m2、m3And m4Refer to for evaluating
The evaluation score that mark is corresponding.
According to the explanation in above-mentioned S41, the present embodiment exist qualified, domestic general, domestic medium, domestic in Comment gathers
5 grades advanced, domestically leading, then in precious Comment gathers, the evaluation system of 5 grades is as shown in table 6:
Table 6
| Sequence number | Enterprise identifies score | Enterprise technology grade |
| 1 | 80-100 | Domestically leading |
| 2 | 60-80 | Domestic advanced person |
| 3 | 40-60 | Domestic medium |
| 4 | 20-40 | Domestic typically |
| 5 | 0-20 | Qualified |
Valuation of enterprise result can be calculated by equation 2 below 0:
Formula 20
G=(M*VT) * 100, G ∈ [0,1]；
Wherein, G is that enterprise identifies score, and M is fuzzy evaluation results, VTFor degree of membership score matrix.
Further, in the present embodiment, above-mentioned degree of membership score matrix can divide and enterprise with specific reference to enterprise in upper table 6 with identifying
The corresponding relation of industry industrial grade determines.
Such as, finally remembered that by above method the G calculated is 79.87, then can be according to enterprise's skill corresponding in table 6 for G
Art grade show that this valuation of enterprise result is for international advanced.
Embodiment two
In the present embodiment, data digging system based on entropy weight algorithm Yu analytic hierarchy process (AHP), specifically includes: achievement data is adopted
Collection terminal, appraisement system build server, data processing server and valuation of enterprise server.
Wherein, achievement data acquisition terminal, gather the initial data for valuation of enterprise, in order to according to the original number gathered
According to carrying out the valuation of enterprise subsequently through native system.
Appraisement system build server, for build with science and technology in enterprise level, enterprise competitiveness, business economic be worth and
Corporate social is worth four aspects and sets up valuation of enterprise system, and determines evaluation index.
Data processing server, for carrying out evaluation index by entropy weight algorithm and analytic hierarchy process (AHP) according to initial data
Process, in order to obtain the index weights of evaluation index, and data processing server and achievement data acquisition terminal and appraisement system
Structure server connects, in order to receive the initial data that achievement data acquisition terminal collects, and opera appraisement system builds clothes
The evaluation index that business device determines processes.
Further, the most concrete including in data processing server: entropy weight algorithm process unit, analytic hierarchy process (AHP) process
Unit and index weights computing unit.
Wherein, entropy weight algorithm process unit, for by the entropy weight of entropy weight algorithm Calculation Estimation index.
Analytic hierarchy process (AHP) processing unit, for by the layer rank weight of analytic hierarchy process (AHP) Calculation Estimation index.
Index weights computing unit, for the entropy weight according to index weights and layer rank weight calculation index weights.
Further, entropy weight algorithm process unit and analytic hierarchy process (AHP) processing unit connect achievement data collection eventually respectively
End, in order to be respectively directed to receive the data of the evaluation index that achievement data acquisition terminal collects simultaneously, and carry out commenting simultaneously
The entropy weight of valency index and the calculating of layer rank weight, in order to the effective treatment effeciency improving valuation of enterprise system.
Concrete, entropy weight algorithm process unit also can further include: normalized subelement, entropy definition subelement
With entropy weight computation subunit.
Wherein, normalized subelement is for carrying out matrix normalization process to evaluation index in initial data；Entropy is fixed
Foster son's unit is for carrying out entropy definition to the evaluation index through normalized；Entropy weight computation subunit is for referring to according to evaluation
Target entropy determines the entropy weight of evaluation index.
Further, normalized subelement, evaluation index is normalized by by the following method:
If total m evaluation object, n evaluation index, then raw data matrix is A=(aij)m×n, wherein, A is original
Data matrix, aijFor i-th evaluation object, the initial data that jth evaluation index is corresponding；
Raw data matrix after raw data matrix is carried out matrix normalization process, after matrix normalization processes
For R=(rij)m×n, wherein, R is the raw data matrix after matrix normalization processes, rijFor i-th evaluation object, jth is commented
The normalization data that valency index is corresponding；
When evaluation index is forward index, evaluation index normalization formula is:
When evaluation index is negative sense index, evaluation index normalization formula is:
Entropy definition subelement, it carries out entropy definition by the following method to evaluation index:
If total m evaluation object, n evaluation index, wherein the formula of the entropy of jth evaluation index is:
Wherein, hjFor i-th evaluation object, the entropy of jth evaluation index, k is the standardized data of m evaluation object,
fijFor i-th evaluation object, the normalization data ratio that jth evaluation index is corresponding；
And,
Entropy weight computation subunit, the entropy weight of its Calculation Estimation index by the following method:
If total m evaluation object, n evaluation index, wherein the formula of the entropy weight of jth evaluation index is:
Wherein, wjEntropy weight for jth evaluation index.
And, analytic hierarchy process (AHP) processing unit also can further include: evaluation index Recurison order hierarchy system construction list
Unit, judgment matrix build subelement, consistency detection subelement, same order weight calculation subelement and layer rank weight calculation sub-list
Unit.
Wherein, evaluation index Recurison order hierarchy system construction subelement is for the layer rank relation according to evaluation index, and foundation is commented
Valency index Recurison order hierarchy system；Judgment matrix builds subelement for being positioned at same single order according in evaluation index Recurison order hierarchy system
The evaluation index of layer, sets up evaluation index multilevel iudge matrix two-by-two, and determines the relatively heavy of evaluation index according to judgment matrix
The property wanted；Consistency detection subelement is for the concordance of test and judge matrix；Same order weight calculation subelement is for for correspondence
In the low stratum evaluation index of ibid stratum's evaluation index, according to the relative importance of evaluation index, evaluation index is carried out
Mode of Level Simple Sequence, determines the same order weight of evaluation index；Layer rank weight calculation subelement is for according to low stratum evaluation index
The same order weight of upper stratum's evaluation index of same order weight and correspondence thereof, carries out total hierarchial sorting to evaluation index, determines and comment
The layer rank weight of valency index.
Further, same order weight calculation subelement, the same order weight of its Calculation Estimation index by the following method:
For judgment matrix B, calculate BW=λmaxThe characteristic root of W and characteristic vector；
To the characteristic vector normalization process calculated, obtain the eigenvectors matrix W=[w after normalization1, w2...
wn,]r, and using the characteristic vector after normalization as the same order weight of evaluation index；
Wherein, W is the eigenvectors matrix after normalization, λmaxIt is characterized root, wnBe feature after the n-th normalization to
Amount, r is characterized moment of a vector rank of matrix.
Further, index weights computing unit connects entropy weight algorithm process unit respectively and analytic hierarchy process (AHP) processes single
Unit, in order to receive entropy weight and the layer of the evaluation index calculated through entropy weight algorithm process unit and analytic hierarchy process (AHP) processing unit
Rank weight is gone forward side by side the calculating of row index weight.
Further, index weights computing unit, the index weights of its Calculation Estimation index by the following method:
Index weights=entropy weight × weight × 0.5,0.5+ layer rank.
Valuation of enterprise server, according to the reference evaluating data of evaluation index set in advance, in conjunction with corresponding index power
Weight, calculates valuation of enterprise data, and this enterprise's par server is connected with data processing server.
Further, what valuation of enterprise server can be concrete includes: appraisement system construction unit, fuzzy algorithmic approach process single
Unit, evaluation result unit and display unit.
Wherein, appraisement system construction unit, it is used for building valuation of enterprise system, and the index set of correspondence and Comment gathers,
Valuation of enterprise system can be determined according to the demand of enterprise customer by this appraisement system construction unit, and comment according to this enterprise
Valency system determines suitable index set and Comment gathers.
Fuzzy algorithmic approach processing unit, for by fuzzy overall evaluation algorithm according to the index weights of evaluation index and reference
Evaluating data calculates valuation of enterprise data.
Evaluation result unit, for according to Comment gathers, it is thus achieved that the valuation of enterprise result that valuation of enterprise data are corresponding, and evaluates
Result unit connects appraisement system construction unit and fuzzy algorithmic approach processing unit respectively, so as to receive appraisement system structure respectively
Build the data message in unit and fuzzy algorithmic approach processing unit, be used for the acquisition of evaluation result with this.
Display unit, is used for showing output valuation of enterprise result, and this display unit connects evaluation result unit, in order to be logical
Cross evaluation result unit and obtain valuation of enterprise result.
Further, valuation of enterprise server may also include memory element, be used for storing in enterprise's par server
The valuation of enterprise system built and the information such as the index set of correspondence, Comment gathers thereof, and this memory element connects appraisement system respectively
Construction unit and evaluation result unit.
The specific works flow process of the present embodiment valuation of enterprise system refers to the detailed description of the above embodiments one.
Embodiment three
Hereinafter, by above-mentioned data digging method based on entropy weight algorithm and analytic hierarchy process (AHP) and system for certain enterprise
Carry out valuation of enterprise.
In the present embodiment, according to the evaluation index of valuation of enterprise, the corresponding relation between each evaluation index, the enterprise of foundation is commented
Valency system.
According to the corresponding relation of each evaluation index in the valuation of enterprise system set up, use entropy weight algorithm and chromatographic assays
Determining the index weights of each evaluation index, the evaluation index and the index weights thereof that finally draw are as shown in table 7:
Table 7
For the valuation of enterprise system of the present embodiment, in long-pending valuation of enterprise system, corresponding each evaluation index is chosen specially
Enterprise of family is with reference to Evaluations matrix and as follows with reference to Evaluations matrix:
Using fuzzy overall evaluation algorithm to be evaluated two-level index, its detailed process is as follows:
According to the content of upper table 7, understanding first two-level index corresponding to first class index according to formula 15 is:
{Q1,Q2,Q3}={ 0.250,0.417,0.333}；
According to formula 17, two-level index carried out the fuzzy evaluating matrix that fuzzy overall evaluation algorithm calculates as follows:
Fuzzy evaluating matrix R can be tried to achieve by above formula1To should look like this:
R1={ 0,0.14575,0.39575,0.4585}
In like manner, according to above method, remaining four first class index are evaluated available R2,R3,R4,R5Fuzzy
Evaluations matrix is as follows:
R2={ 0.08325,0.2665,0.35,0.30025}
R3={ 0.09725,0.34725,0.40275,0.15275}
R4={ 0.08325,0.26375,0.34725,0.30575}
R5={ 0,0.1875,0.4375,0.375}
According to the fuzzy evaluating matrix of the two-level index calculated above, then first order calculation index, by enterprise is carried out
Evaluate.
The matrix W of index weights={ p1,p2,p3,p4,p5, the index weights according to upper table 7 record can be by above-mentioned finger
Mark weight matrix is specifically written as:
W={0.389,0.262,0.107,0.18,0.062}
Further, in conjunction with this index weights matrix, can further enterprise be evaluated.
That is, index weights matrix W is carried out fuzzy operation with fuzzy evaluating matrix R, show that fuzzy evaluation results is:
Above formula is carried out abbreviation calculating, obtains equation below:
M=W × R={m1,m2,m3,m4}={ 0.0472,0.2227,0.3783,0.3516}
According to the valuation of enterprise system set up, set up Comment gathers.
Concrete, in the present embodiment, for this project the evaluation of Enterprise Integrated quality is divided into poor, in, good, excellent 4 etc.
Level, is designated as
Corresponding degree of membership score matrix is:
J={0.25,0.5,0.75,1}
Based on above-mentioned content, show according to comprehensive evaluation result, this enterprise four degree of membership differences, in, good, excellent ratio
Example is respectively as follows: m1,m2,m3,m4。
Then according to below equation can draw final enterprise identify score:
G=M*JT={ 0.0472,0.2227,0.3783,0.3516}*{0.25,0.5,0.75,1}T× 100=75.86
Can draw according to above evaluation procedure, valuation of enterprise score G of this enterprise is 75.86, and the value of G determines enterprise
Overall qualities, this enterprise is in excellent level.
The data digging method based on entropy weight algorithm and analytic hierarchy process (AHP) of the present invention and system, omnibearing by setting up
Valuation of enterprise system, technical advance and practical value to enterprise are evaluated, simultaneously by entropy weight algorithm and step analysis
Method and the combination of Field Using Fuzzy Comprehensive Assessment so that valuation of enterprise result is more objective, science, in order to show more accurately
The power of science and technology in enterprise ability, formulates interim strategic objective for enterprise and provides important references.
Last it is noted that above example is only in order to illustrate technical scheme, it is not intended to limit；Although
With reference to previous embodiment, the present invention is described in detail, it will be understood by those within the art that: it still may be used
So that the technical scheme described in foregoing embodiments to be modified, or wherein portion of techniques feature is carried out equivalent；
And these amendment or replace, do not make appropriate technical solution essence depart from various embodiments of the present invention technical scheme spirit and
Scope.
Claims (19)
1. a data digging method based on entropy weight algorithm Yu analytic hierarchy process (AHP), it is characterised in that comprise the following steps:
S1, gather for the initial data of valuation of enterprise；
S2, with science and technology in enterprise level, enterprise competitiveness, business economic be worth and corporate social be worth four aspects set up enterprise
Appraisement system, and determine evaluation index；
S3, employing entropy weight algorithm and analytic hierarchy process (AHP) determine the index weights of each evaluation index in valuation of enterprise；
S4, use according to the index weights of standard diagrams and correspondence thereof fuzzy overall evaluation algorithm to carry out valuation of enterprise, looked forward to
Industry evaluation result.
Data digging method based on entropy weight algorithm Yu analytic hierarchy process (AHP) the most according to claim 1, it is characterised in that
In described S3, entropy weight algorithm comprises the following steps:
S31, evaluation index in initial data is carried out matrix normalization process；
S32, the evaluation index processed through matrix normalization is carried out entropy definition；
S33, entropy according to evaluation index determine the entropy weight of evaluation index；
The middle-level analytic process of described S2 comprises the following steps:
S34, layer rank relation according to evaluation index, set up evaluation index Recurison order hierarchy system；
S35, according to evaluation index Recurison order hierarchy system is positioned at the evaluation index of same stratum, set up evaluation index and compare two-by-two
Judgment matrix, and the relative importance of evaluation index is determined according to judgment matrix；
S36, for the low stratum evaluation index corresponding to ibid stratum's evaluation index, relatively important according to evaluation index
Property, evaluation index is carried out Mode of Level Simple Sequence, determines the same order weight of evaluation index；
S37, according to the same order weight of low stratum evaluation index and the same order weight of upper stratum's evaluation index of correspondence thereof, to commenting
Valency index carries out total hierarchial sorting, determines the layer rank weight of evaluation index；
Entropy weight according to evaluation index and layer rank weight, calculate described index weights by linear weighted function averaging method.
Data digging method based on entropy weight algorithm Yu analytic hierarchy process (AHP) the most according to claim 2, it is characterised in that
In described S31, the matrix normalization of evaluation index processes and calculates by the following method:
S311, setting total m evaluation object, n evaluation index, then raw data matrix is A=(aij)m×n, wherein, A is original
Data matrix, aijFor i-th evaluation object, the initial data that jth evaluation index is corresponding；
S312, raw data matrix is carried out matrix normalization process after, through matrix normalization process after initial data square
Battle array is R=(rij)m×n, wherein, R is the raw data matrix after matrix normalization processes, rijFor i-th evaluation object, jth
The normalization data that evaluation index is corresponding；
S313, when evaluation index is forward index, evaluation index normalization formula is:
When evaluation index is negative sense index, evaluation index normalization formula is:
Data digging method based on entropy weight algorithm Yu analytic hierarchy process (AHP) the most according to claim 2, it is characterised in that
The evaluation index processed through matrix normalization is carried out entropy definition by described S32 calculate by the following method:
If total m evaluation object, n evaluation index, wherein the formula of the entropy of jth evaluation index is:
Wherein, hjFor i-th evaluation object, the entropy of jth evaluation index, k is the standardized data of m evaluation object, fijFor
I-th evaluation object, the normalization data ratio that jth evaluation index is corresponding；
And,
Data digging method based on entropy weight algorithm Yu analytic hierarchy process (AHP) the most according to claim 2, it is characterised in that
In described S33, the entropy weight of evaluation index calculates by the following method:
If total m evaluation object, n evaluation index, wherein the formula of the entropy weight of jth evaluation index is:
Wherein, wjEntropy weight for jth evaluation index.
Data digging method based on entropy weight algorithm Yu analytic hierarchy process (AHP) the most according to claim 2, it is characterised in that
In described S36 before carrying out Mode of Level Simple Sequence, the concordance of test and judge matrix；
When judging that judgment matrix meets concordance, then evaluation index is carried out Mode of Level Simple Sequence；
When judging that judgment matrix is unsatisfactory for concordance, then adjust the judgment matrix of correspondence.
Data digging method based on entropy weight algorithm Yu analytic hierarchy process (AHP) the most according to claim 2, it is characterised in that
According to Mode of Level Simple Sequence, described S36 determines that same order weight calculates by the following method:
For judgment matrix B, calculate BW=λmaxThe characteristic root of W and characteristic vector；
To the characteristic vector normalization process calculated, obtain the eigenvectors matrix W=[w after normalization1, w2... wn,]r,
And using the characteristic vector after this normalization as the same order weight of evaluation index；
Wherein, W is the eigenvectors matrix after normalization, λmaxIt is characterized root, wnBeing the characteristic vector after the n-th normalization, r is
The order of eigenvectors matrix.
Data digging method based on entropy weight algorithm Yu analytic hierarchy process (AHP) the most according to claim 2, it is characterised in that
Described linear weighted function averaging method is: index weights=entropy weight × weight × 0.5,0.5+ layer rank.
Data digging method based on entropy weight algorithm Yu analytic hierarchy process (AHP) the most according to claim 1, it is characterised in that
In described S4, index weights according to standard diagrams and correspondence thereof carries out valuation of enterprise and comprises the following steps:
S41, according to valuation of enterprise Establishing business evaluation indicator collection and Comment gathers, and determine reference according to expert opinion data
Evaluations matrix；
S42, according to index set and with reference to Evaluations matrix use fuzzy overall evaluation algorithm evaluation index is evaluated, it is thus achieved that mould
Stick with paste Evaluations matrix；
S43, draw fuzzy evaluation results according to the index weights of evaluation index and fuzzy evaluating matrix, and tie according to fuzzy evaluation
The comment that fruit is corresponding in Comment gathers determines valuation of enterprise result.
10. a data digging system based on entropy weight algorithm Yu analytic hierarchy process (AHP), it is characterised in that including:
Achievement data acquisition terminal, it gathers the initial data for valuation of enterprise；
Appraisement system builds server, and it is worth and enterprise with science and technology in enterprise level, enterprise competitiveness, business economic for building
Four aspects of industry social value set up valuation of enterprise system, and determine evaluation index；
Data processing server, it builds server with described achievement data acquisition terminal and appraisement system respectively and is connected, according to
Evaluation index is processed by initial data by entropy weight algorithm and analytic hierarchy process (AHP), in order to evaluation index index weigh
Weight；
Valuation of enterprise server, it is connected with described data processing server, comments according to the reference of evaluation index set in advance
Valence mumber evidence, in conjunction with corresponding index weights, calculates valuation of enterprise data.
11. data digging systems based on entropy weight algorithm Yu analytic hierarchy process (AHP) according to claim 10, it is characterised in that
Described data processing server includes:
Entropy weight algorithm process unit, it is for by the entropy weight of entropy weight algorithm Calculation Estimation index；
Analytic hierarchy process (AHP) processing unit, it is for by the layer rank weight of analytic hierarchy process (AHP) Calculation Estimation index；
Index weights computing unit, it is for the entropy weight according to index weights and layer rank weight calculation index weights.
12. data digging systems based on entropy weight algorithm Yu analytic hierarchy process (AHP) according to claim 11, it is characterised in that
Described entropy weight algorithm process unit includes:
Normalized subelement, it is for carrying out matrix normalization process to evaluation index in initial data；
Entropy definition subelement, it is for carrying out entropy definition to the evaluation index through normalized；
Entropy weight computation subunit, it for determining the entropy weight of evaluation index according to the entropy of evaluation index.
13. data digging systems based on entropy weight algorithm Yu analytic hierarchy process (AHP) according to claim 12, it is characterised in that
Described normalized subelement, evaluation index is normalized by by the following method:
If total m evaluation object, n evaluation index, then raw data matrix is A=(aij)m×n, wherein, A is initial data
Matrix, aijFor i-th evaluation object, the initial data that jth evaluation index is corresponding；
After raw data matrix is carried out matrix normalization process, the raw data matrix after matrix normalization processes is R
=(rij)m×n, wherein, R is the raw data matrix after matrix normalization processes, rijFor i-th evaluation object, jth evaluation
The normalization data that index is corresponding；
When evaluation index is forward index, evaluation index normalization formula is:
When evaluation index is negative sense index, evaluation index normalization formula is:
14. data digging systems based on entropy weight algorithm Yu analytic hierarchy process (AHP) according to claim 13, it is characterised in that
Described entropy definition subelement, it carries out entropy definition by the following method to evaluation index:
If total m evaluation object, n evaluation index, wherein the formula of the entropy of jth evaluation index is:
Wherein, hjFor i-th evaluation object, the entropy of jth evaluation index, k is the standardized data of m evaluation object, fijFor
I-th evaluation object, the normalization data ratio that jth evaluation index is corresponding；
And,
15. data digging systems based on entropy weight algorithm Yu analytic hierarchy process (AHP) according to claim 14, it is characterised in that
Described entropy weight computation subunit, the entropy weight of its Calculation Estimation index by the following method:
If total m evaluation object, n evaluation index, wherein the formula of the entropy weight of jth evaluation index is:
Wherein, wjEntropy weight for jth evaluation index.
16. data digging systems based on entropy weight algorithm Yu analytic hierarchy process (AHP) according to claim 11, it is characterised in that
Described analytic hierarchy process (AHP) processing unit includes:
Evaluation index Recurison order hierarchy system construction subelement, it, for the layer rank relation according to evaluation index, sets up evaluation index
Recurison order hierarchy system；
Judgment matrix builds subelement, and it is for referring to according to the evaluation being positioned at same stratum in evaluation index Recurison order hierarchy system
Mark, sets up evaluation index multilevel iudge matrix two-by-two, and determines the relative importance of evaluation index according to judgment matrix；
Consistency detection subelement, it is for the concordance of test and judge matrix；
Same order weight calculation subelement, it is for for the low stratum evaluation index corresponding to ibid stratum's evaluation index, root
According to the relative importance of evaluation index, evaluation index is carried out Mode of Level Simple Sequence, determine the same order weight of evaluation index；
Layer rank weight calculation subelement, it is for commenting according to the same order weight of low stratum evaluation index and a upper stratum of correspondence thereof
The same order weight of valency index, carries out total hierarchial sorting to evaluation index, determines the layer rank weight of evaluation index.
17. data digging systems based on entropy weight algorithm Yu analytic hierarchy process (AHP) according to claim 16, it is characterised in that
Described same order weight calculation subelement, the same order weight of its Calculation Estimation index by the following method:
For judgment matrix B, calculate BW=λmaxThe characteristic root of W and characteristic vector；
To the characteristic vector normalization process calculated, obtain the eigenvectors matrix W=[w after normalization1, w2... wn,]r,
And using the characteristic vector after normalization as the same order weight of evaluation index；
Wherein, W is the eigenvectors matrix after normalization, λmaxIt is characterized root, wnBeing the characteristic vector after the n-th normalization, r is
The order of eigenvectors matrix.
18. data digging systems based on entropy weight algorithm Yu analytic hierarchy process (AHP) according to claim 17, it is characterised in that
Described index weights computing unit, the index weights of its Calculation Estimation index by the following method:
Index weights=entropy weight × weight × 0.5,0.5+ layer rank.
19. data digging systems based on entropy weight algorithm Yu analytic hierarchy process (AHP) according to claim 10, it is characterised in that
Described valuation of enterprise server includes:
Appraisement system construction unit, it is used for building valuation of enterprise system, and the index set of correspondence and Comment gathers；
Fuzzy algorithmic approach processing unit, it is for commenting according to the index weights of evaluation index and reference by fuzzy overall evaluation algorithm
Valence mumber is according to calculating valuation of enterprise data；
Evaluation result unit, it is for according to Comment gathers, it is thus achieved that the valuation of enterprise result that valuation of enterprise data are corresponding；
Display unit, it is used for showing output valuation of enterprise result.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201610791303.9A CN106326473A (en) | 2016-08-31 | 2016-08-31 | Data mining method based on entropy weight algorithm and analytic hierarchy process and system thereof |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201610791303.9A CN106326473A (en) | 2016-08-31 | 2016-08-31 | Data mining method based on entropy weight algorithm and analytic hierarchy process and system thereof |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN106326473A true CN106326473A (en) | 2017-01-11 |
Family
ID=57790085
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201610791303.9A Pending CN106326473A (en) | 2016-08-31 | 2016-08-31 | Data mining method based on entropy weight algorithm and analytic hierarchy process and system thereof |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN106326473A (en) |
Cited By (14)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106875047A (en) * | 2017-01-23 | 2017-06-20 | 国网湖南省电力公司 | Reservoir watershed Runoff Forecast method and system |
| CN107609754A (en) * | 2017-08-25 | 2018-01-19 | 国网辽宁省电力有限公司 | A kind of grid information system performance monitoring system and method |
| CN108053094A (en) * | 2017-11-15 | 2018-05-18 | 九次方大数据信息集团有限公司 | A kind of weight grade evaluation method and system |
| CN108629023A (en) * | 2018-05-09 | 2018-10-09 | 北京京东金融科技控股有限公司 | Data digging method, device and computer readable storage medium |
| CN108898275A (en) * | 2018-05-30 | 2018-11-27 | 北京农业信息技术研究中心 | A kind of milk cattle cultivating environmental degree of comfort evaluation method and system based on cloud model |
| CN108898890A (en) * | 2018-08-29 | 2018-11-27 | 中国民用航空总局第二研究所 | Blank pipe operational efficiency grade appraisal procedure and its device |
| CN108921438A (en) * | 2018-07-10 | 2018-11-30 | 国网福建省电力有限公司 | A kind of power distribution network regulation and administration weak link identification method based on cascade weight |
| CN109002969A (en) * | 2018-06-27 | 2018-12-14 | 中国水利水电科学研究院 | A kind of evaluation method of regional water use competitiveness |
| CN109102183A (en) * | 2018-08-01 | 2018-12-28 | 北京会达分享信息技术有限公司 | A kind of selection method and device of the corporate resources shared |
| CN109189821A (en) * | 2018-08-01 | 2019-01-11 | 成都数联铭品科技有限公司 | Data analysis system based on affiliated party's index system |
| CN111080158A (en) * | 2019-12-26 | 2020-04-28 | 安徽揣菲克科技有限公司 | Urban intersection traffic danger index evaluation method based on composite weight |
| CN113256075A (en) * | 2021-04-29 | 2021-08-13 | 浙江非线数联科技股份有限公司 | Enterprise risk level evaluation method based on hierarchical analysis and fuzzy comprehensive evaluation method |
| CN113762795A (en) * | 2021-09-13 | 2021-12-07 | 浙江万维空间信息技术有限公司 | Industrial chain diagnosis method and system based on hierarchical analysis |
| CN115146939A (en) * | 2022-06-24 | 2022-10-04 | 国网江苏省电力有限公司经济技术研究院 | Power grid engineering project comprehensive technical level pre-evaluation method |
-
2016
- 2016-08-31 CN CN201610791303.9A patent/CN106326473A/en active Pending
Cited By (17)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106875047B (en) * | 2017-01-23 | 2021-03-16 | 国网湖南省电力公司 | Reservoir basin runoff forecasting method and system |
| CN106875047A (en) * | 2017-01-23 | 2017-06-20 | 国网湖南省电力公司 | Reservoir watershed Runoff Forecast method and system |
| CN107609754A (en) * | 2017-08-25 | 2018-01-19 | 国网辽宁省电力有限公司 | A kind of grid information system performance monitoring system and method |
| CN108053094A (en) * | 2017-11-15 | 2018-05-18 | 九次方大数据信息集团有限公司 | A kind of weight grade evaluation method and system |
| CN108629023A (en) * | 2018-05-09 | 2018-10-09 | 北京京东金融科技控股有限公司 | Data digging method, device and computer readable storage medium |
| CN108898275A (en) * | 2018-05-30 | 2018-11-27 | 北京农业信息技术研究中心 | A kind of milk cattle cultivating environmental degree of comfort evaluation method and system based on cloud model |
| CN108898275B (en) * | 2018-05-30 | 2020-12-29 | 北京农业信息技术研究中心 | Cloud model-based cow breeding environment comfort evaluation method and system |
| CN109002969A (en) * | 2018-06-27 | 2018-12-14 | 中国水利水电科学研究院 | A kind of evaluation method of regional water use competitiveness |
| CN108921438A (en) * | 2018-07-10 | 2018-11-30 | 国网福建省电力有限公司 | A kind of power distribution network regulation and administration weak link identification method based on cascade weight |
| CN108921438B (en) * | 2018-07-10 | 2022-03-08 | 国网福建省电力有限公司 | Power distribution network regulation and control management weak link identification method based on cascade weight |
| CN109102183A (en) * | 2018-08-01 | 2018-12-28 | 北京会达分享信息技术有限公司 | A kind of selection method and device of the corporate resources shared |
| CN109189821A (en) * | 2018-08-01 | 2019-01-11 | 成都数联铭品科技有限公司 | Data analysis system based on affiliated party's index system |
| CN108898890A (en) * | 2018-08-29 | 2018-11-27 | 中国民用航空总局第二研究所 | Blank pipe operational efficiency grade appraisal procedure and its device |
| CN111080158A (en) * | 2019-12-26 | 2020-04-28 | 安徽揣菲克科技有限公司 | Urban intersection traffic danger index evaluation method based on composite weight |
| CN113256075A (en) * | 2021-04-29 | 2021-08-13 | 浙江非线数联科技股份有限公司 | Enterprise risk level evaluation method based on hierarchical analysis and fuzzy comprehensive evaluation method |
| CN113762795A (en) * | 2021-09-13 | 2021-12-07 | 浙江万维空间信息技术有限公司 | Industrial chain diagnosis method and system based on hierarchical analysis |
| CN115146939A (en) * | 2022-06-24 | 2022-10-04 | 国网江苏省电力有限公司经济技术研究院 | Power grid engineering project comprehensive technical level pre-evaluation method |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN106326473A (en) | Data mining method based on entropy weight algorithm and analytic hierarchy process and system thereof | |
| Zhou et al. | Type-2 fuzzy multi-objective DEA model: An application to sustainable supplier evaluation | |
| Abdullah et al. | A new type-2 fuzzy set of linguistic variables for the fuzzy analytic hierarchy process | |
| Silva et al. | Network structure analysis of the Brazilian interbank market | |
| CN110110981B (en) | Credit rating default probability measure and risk early warning method | |
| Lu et al. | Bank discrimination, holding bank ownership, and economic consequences: Evidence from China | |
| Galagedera et al. | Modeling leakage in two-stage DEA models: An application to US mutual fund families | |
| CN106372952A (en) | Objective and subjective weight determining multi-model compositional verification-based enterprise credit assessment method and system | |
| CN107633265A (en) | For optimizing the data processing method and device of credit evaluation model | |
| CN107437151A (en) | Enterprise credit risk method | |
| CN108537435A (en) | Performance of banking industry evaluation method, system, equipment, storage medium | |
| Yun et al. | Analysis of the teaching quality on deep learning-based innovative ideological political education platform | |
| CN110276556A (en) | One kind being based on shifty environmental carrying capacity index weights calculation method and device | |
| CN109345050A (en) | A kind of quantization transaction prediction technique, device and equipment | |
| CN108573339A (en) | A kind of consumer's net purchase methods of risk assessment of multi objective Project Decision Method | |
| CN105787269A (en) | Heterogeneous multi-attribute variable-weight decision-making method based on regret theory | |
| Labijak-Kowalska et al. | Experimental comparison of results provided by ranking methods in Data Envelopment Analysis | |
| CN106600116A (en) | Sludge plate frame dewatering effect evaluation method based on FAHP method | |
| Ebrahimi et al. | A novel mixed binary linear DEA model for ranking decision-making units with preference information | |
| Wanke et al. | Revisiting camels rating system and the performance of Asean banks: a comprehensive mcdm/z-numbers approach | |
| Ohene-Asare et al. | Banking efficiency analysis under corporate social responsibilities | |
| Sajjad et al. | Investigating the impact of external debt on economic growth: a case study of Pakistan | |
| CN105303302A (en) | Power grid evaluating indicator correlation analysis method, apparatus and computing apparatus | |
| CN105447117A (en) | User clustering method and apparatus | |
| Perera et al. | On the cross‐methodological validation of bank efficiency assessments |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20170111 |
|
| RJ01 | Rejection of invention patent application after publication |