CN109376966A - A kind of Optimization Method of Index System and device - Google Patents
A kind of Optimization Method of Index System and device Download PDFInfo
- Publication number
- CN109376966A CN109376966A CN201811528291.6A CN201811528291A CN109376966A CN 109376966 A CN109376966 A CN 109376966A CN 201811528291 A CN201811528291 A CN 201811528291A CN 109376966 A CN109376966 A CN 109376966A
- Authority
- CN
- China
- Prior art keywords
- index
- cluster
- representative
- association
- class
- 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
- 238000000034 method Methods 0.000 title claims abstract description 63
- 238000005457 optimization Methods 0.000 title claims abstract description 36
- 238000012545 processing Methods 0.000 claims abstract description 31
- 238000012216 screening Methods 0.000 claims abstract description 14
- 238000011156 evaluation Methods 0.000 claims abstract description 10
- 230000008569 process Effects 0.000 abstract description 11
- 230000004927 fusion Effects 0.000 abstract description 3
- 239000010755 BS 2869 Class G Substances 0.000 description 7
- 238000010219 correlation analysis Methods 0.000 description 6
- 238000010586 diagram Methods 0.000 description 6
- 239000011159 matrix material Substances 0.000 description 6
- 230000000875 corresponding effect Effects 0.000 description 4
- 238000013528 artificial neural network Methods 0.000 description 3
- 238000005259 measurement Methods 0.000 description 3
- 238000010606 normalization Methods 0.000 description 3
- 230000009466 transformation Effects 0.000 description 3
- 230000007547 defect Effects 0.000 description 2
- 238000005516 engineering process Methods 0.000 description 2
- 230000005484 gravity Effects 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- PEDCQBHIVMGVHV-UHFFFAOYSA-N Glycerine Chemical compound OCC(O)CO PEDCQBHIVMGVHV-UHFFFAOYSA-N 0.000 description 1
- 238000004458 analytical method Methods 0.000 description 1
- 230000008901 benefit Effects 0.000 description 1
- 230000015572 biosynthetic process Effects 0.000 description 1
- 238000004364 calculation method Methods 0.000 description 1
- 230000008859 change Effects 0.000 description 1
- 238000007621 cluster analysis Methods 0.000 description 1
- 230000002596 correlated effect Effects 0.000 description 1
- 238000009826 distribution Methods 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 238000001914 filtration Methods 0.000 description 1
- 238000007781 pre-processing Methods 0.000 description 1
- 238000002360 preparation method Methods 0.000 description 1
- 230000000750 progressive effect Effects 0.000 description 1
- 230000009467 reduction Effects 0.000 description 1
- 238000003860 storage Methods 0.000 description 1
- 230000009885 systemic effect Effects 0.000 description 1
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
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/23—Clustering techniques
-
- 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)
- Economics (AREA)
- Theoretical Computer Science (AREA)
- Strategic Management (AREA)
- Development Economics (AREA)
- Data Mining & Analysis (AREA)
- Entrepreneurship & Innovation (AREA)
- General Physics & Mathematics (AREA)
- Physics & Mathematics (AREA)
- Tourism & Hospitality (AREA)
- Quality & Reliability (AREA)
- Operations Research (AREA)
- General Business, Economics & Management (AREA)
- Marketing (AREA)
- Game Theory and Decision Science (AREA)
- Educational Administration (AREA)
- Life Sciences & Earth Sciences (AREA)
- Artificial Intelligence (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Bioinformatics & Computational Biology (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Evolutionary Biology (AREA)
- Evolutionary Computation (AREA)
- General Engineering & Computer Science (AREA)
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
Abstract
This application provides a kind of Optimization Method of Index System and devices, which comprises pre-processes to each index in index system, so that the different each indices non-dimension of dimension in index system；Clustering processing is carried out to pretreated each index, obtains multiple index clusters；Based on predetermined evaluation method, from least one maximum index of representative degree is selected in each index cluster as the representative index in each index cluster；Merge the representative index in each index cluster, the index system after being optimized.It can be seen that, application scheme carries out index screening in cluster and class based on each index to index system, realize index system optimization, the Cluster-Fusion combined influence factor of various aspects, to which the application passes through the combined influence factor of fusion various aspects, index system is optimized, may make the optimization processing of index system with more comprehensive.
Description
Technical field
The invention belongs to the screening of index, preferentially technical field more particularly to a kind of Optimization Method of Index System and device.
Background technique
In actual application environment, the quantity of index is often not The more the better, and excessive index can bring information redundancy, instead
And the accuracy of operation situation assessment is influenced, therefore, it is proposed to the viewpoint of index system optimization.The purpose of index system optimization exists
In the key node for being directed to covering company whole operation flow, screening appropriate is carried out to the existing index system of company, thus right
Company operation situation has one more accurately and efficiently to assess.
In general, the optimization of index system has the formation of a large amount of index reduced often by statistical check means
One group of significant statistical property finds out correlated variables for linear problem and some specific nonlinear problems, rejects correlation
High and little discrimination index.Existing Optimization Method of Index System mainly includes neural network, rough set attribute letter
Provisional constitution and Vague theory method etc., however these methods have the defects that it is certain, by taking neural network as an example, though neural network
So there is stronger generalization ability, but need artificial determination core index as output, with core index for main mesh
Mark is necessarily required to the selection biggish target variable of influence degree, there is certain limitation；In consideration of it, this field need to provide one kind preferably
Index system optimization scheme, at least overcome existing Optimization Method of Index System certain aspect defect.
Summary of the invention
In view of this, the purpose of the present invention is to provide a kind of Optimization Method of Index System and devices, it is intended to pass through fusion
The combined influence factor of various aspects, optimizes index system, so that the optimization processing of index system is with comprehensive.
For this purpose, the present invention is disclosed directly below technical solution:
A kind of Optimization Method of Index System, comprising:
Each index in index system is pre-processed, so that the different each index of dimension in index system
Nondimensionalization；
Clustering processing is carried out to pretreated each index, obtains multiple index clusters；
Based on predetermined evaluation method, at least one maximum index of representative degree is selected as institute from each index cluster
State the representative index in each index cluster；
Merge the representative index in each index cluster, the index system after being optimized.
The above method, it is preferred that each index in index system pre-processes, so that measuring in index system
The different each indices non-dimension of guiding principle, comprising:
Each index in index system is grouped based on pre-defined rule, obtains multiple index groups；
Scheduled nondimensionalization processing is carried out to the index in each index group.
The above method, it is preferred that it is described that each index in index system is grouped based on pre-defined rule, it obtains more
A index group, comprising:
Each index in index system is grouped by the value type of index, obtains multiple index groups；
The value type of the index includes for indicating the first kind that index value is the bigger the better, for indicating index value
The smaller the better Second Type, for indicating the index value third type better closer to a certain predetermined value.
The above method, it is preferred that it is described that clustering processing is carried out to pretreated each index, it is poly- to obtain multiple indexs
Class, comprising:
Multiple groups based on index each in index system observe data, calculate every two in pretreated each index and refer to
Multiple grey relational grade numerical value between mark；
According to multiple grey relational grade numerical value between every two index, the average degree of association between every two index is calculated；
Based on the average degree of association between every two index, pretreated each index is carried out using predetermined clusters method
Clustering processing obtains multiple index clusters, wherein the average degree of association between class between any two index is higher than any two in class
The average degree of association between index.
The above method, it is preferred that it is described to be based on predetermined evaluation method, representative maximum is selected from each index cluster
At least one index as the representative index in each index cluster, comprising:
Observation data based on index calculate the grey relational grade in each index cluster between every two index；
Based on the grey relational grade in each index cluster between every two index, calculate in index cluster each index with
The average degree of association of grey relational grade between other interior each indexs of class, as representative degree of the index in class；
Select generation of at least one the maximum index of representative degree as each index cluster in each index cluster
Table index.
A kind of index system optimization device, comprising:
Pretreatment unit, for being pre-processed to each index in index system, so that dimension is not in index system
Same each indices non-dimension；
Cluster cell obtains multiple index clusters for carrying out clustering processing to pretreated each index；
Screening unit is used to be based on predetermined evaluation method, and it is maximum at least to select representative degree from each index cluster
One index is as the representative index in each index cluster；
Combining unit, the index system for merging the representative index in each index cluster, after being optimized.
Above-mentioned apparatus, it is preferred that the pretreatment unit is specifically used for:
Each index in index system is grouped based on pre-defined rule, obtains multiple index groups；
Scheduled nondimensionalization processing is carried out to the index in each index group.
Above-mentioned apparatus, it is preferred that the pretreatment unit carries out each index in index system based on pre-defined rule
Grouping, obtains multiple index groups, specifically includes:
Each index in index system is grouped by the value type of index, obtains multiple index groups；
The value type of the index includes for indicating the first kind that index value is the bigger the better, for indicating index value
The smaller the better Second Type, for indicating the index value third type better closer to a certain predetermined value.
Above-mentioned apparatus, it is preferred that the cluster cell is specifically used for:
Multiple groups based on index each in index system observe data, calculate every two in pretreated each index and refer to
Multiple grey relational grade numerical value between mark；
According to multiple grey relational grade numerical value between every two index, the average degree of association between every two index is calculated；
Based on the average degree of association between every two index, pretreated each index is carried out using predetermined clusters method
Clustering processing obtains multiple index clusters, wherein the average degree of association between class between any two index is higher than any two in class
The average degree of association between index.
Above-mentioned apparatus, it is preferred that the screening unit is specifically used for:
Observation data based on index calculate the grey relational grade in each index cluster between every two index；
Based on the grey relational grade in each index cluster between every two index, calculate in index cluster each index with
The average degree of association of grey relational grade between other interior each indexs of class, as representative degree of the index in class；
Select generation of at least one the maximum index of representative degree as each index cluster in each index cluster
Table index.
Based on above scheme it is found that this application provides a kind of Optimization Method of Index System and devices, which comprises
Each index in index system is pre-processed, so that the different each index dimensionless of dimension in index system
Change；Clustering processing is carried out to pretreated each index, obtains multiple index clusters；Based on predetermined evaluation method, from each
At least one maximum index of representative degree is selected in index cluster as the representative index in each index cluster；It closes
And the representative index in each index cluster, the index system after being optimized.It can be seen that application scheme is based on to finger
Each index of mark system carries out index screening in cluster and class, realizes index system optimization, Cluster-Fusion various aspects
Combined influence factor, so that the application is optimized index system, can be made by the combined influence factor of fusion various aspects
The optimization processing of index system is obtained with more comprehensive.
Detailed description of the invention
In order to more clearly explain the embodiment of the invention or the technical proposal in the existing technology, to embodiment or will show below
There is attached drawing needed in technical description to be briefly described, it should be apparent that, the accompanying drawings in the following description is only this
The embodiment of invention for those of ordinary skill in the art without creative efforts, can also basis
The attached drawing of offer obtains other attached drawings.
Fig. 1 is the flow diagram of Optimization Method of Index System provided by the embodiments of the present application；
Fig. 2 is the flow diagram provided by the embodiments of the present application that clustering processing is carried out to index；
Fig. 3 is the screening process schematic diagram of representative index in class provided by the embodiments of the present application；
Fig. 4 is the structural schematic diagram of index system optimization device provided by the embodiments of the present application.
Specific embodiment
For the sake of quoting and understanding, hereafter used in technical term, write a Chinese character in simplified form or abridge summary be explained as follows:
Index: i.e. pre- interim plan attended index, specification and standard, are the concepts for illustrating total number feature.Index
It is generally made of index name and index value two parts, it embodies two aspects of regularity of things qualitative definition and amount
The characteristics of.Index can illustrate overall feature, but only react overall quantative attribute.Index mentioned in the present invention is to be used for
Describe and evaluate the measurement standard of certain enterprise or unit traffic-operating period.
Grey correlation analysis: grey correlation analysis is the basic content of gray system theory, its basic thought is basis
Similarity degree between curve carrys out the correlation degree between factor of judgment.Grey correlation analysis belongs to several from its way of thinking
The scope of where reason is compared the geometry for reflecting that the data sequence of each factor variation characteristic is carried out.For measurement factor it
Between correlation degree the degree of association, be exactly as obtained from the comparison of the invariance curve between factor.This method breaches tradition
Accurate mathematical never allows ambiguous constraint, has that principle is simple, be easy to grasp, calculates that easy, sequence is clear, to data
The features such as correlation type between distribution pattern and variable is without particular/special requirement, therefore there is great practical application value.
Hierarchical Clustering: Hierarchical Clustering is by the method for each sample divide into several classes, and basic thought is first that each sample is each
Regard a kind of as, then provide the distance between class and class, selection is merged into new one kind apart from the smallest a pair, calculate new class and
The distance between other classes, then two nearest classes of distance are merged, reduction is a kind of every time in this way, until all samples are combined into one
Until class.
Following will be combined with the drawings in the embodiments of the present invention, and technical solution in the embodiment of the present invention carries out clear, complete
Site preparation description, it is clear that described embodiments are only a part of the embodiments of the present invention, instead of all the embodiments.It is based on
Embodiment in the present invention, it is obtained by those of ordinary skill in the art without making creative efforts every other
Embodiment shall fall within the protection scope of the present invention.
This application provides a kind of Optimization Method of Index System and device, the major technique thinking of this method and device is benefit
It is carried out with the method for the degree of association between each index in Grey Correlation Analysis Theory analysis indexes system, and utilization cluster by the degree of association
Cluster filters out the highest index of representative degree in every class, final merging is each on this basis by screening to index in class
The highest representative degree index filtered out in class is as the index system after optimization.It below will be by specific embodiment to the application
Method and device be described in detail.
It is a kind of flow diagram of Optimization Method of Index System provided by the embodiments of the present application, the present embodiment with reference to Fig. 1
In, described method includes following steps:
Step 101 pre-processes each index in index system, so that different described of dimension in index system
Each indices non-dimension.
The pretreatment that the step carries out index belongs to the normalisation section of data.
The purpose of data normalization: since different indexs usually have different units and different degrees of variation.It is different
Unit often make the practice of coefficient explain occur it is difficult.When variables itself different simultaneously has larger difference, it can make to calculate
Coefficient of relationship in specific gravity shared by different indexs differ greatly.In order to eliminate dimension impact and variable itself variation size logarithm
It is worth the influence of size, therefore needs data normalization, it is described to standardize namely nondimensionalization processing is carried out to data.
This step is primarily based on pre-defined rule in index system when pre-processing to the index in index system
Each index is grouped, and obtains multiple index groups.Wherein, if relevance since index value type is different, between index
Can be relatively small, it is therefore, described that each index in index system is grouped based on pre-defined rule, it can be by index
Value type is grouped each index in index system, for example, by the first kind being the bigger the better for indicating index value
Type, for indicating the smaller the better Second Type of index value, for indicating the index value third better closer to a certain predetermined value
Each index in index system is divided into different groups and handled etc. by type.
If being grouped the latter group includes n index, each index includes m observation data, then specifically has index below
Data sequence:
X1=(x1(1), x1(2) ..., x1(n))
X2=(x2(1), x2(2) ..., x2(n))
Xm=(xm(1), xm(2) ..., xm(n))
On this basis, can continue to carry out nondimensionalization processing to each group index after grouping.
Wherein, for the index in same group, nondimensionalization below can be carried out to index according to the concrete condition of index
Processing:
(1) index is the bigger the better type
The type index for example can be critical workflow index in practice.
Standardized method are as follows:
Former index is bigger, and new index is bigger, generates new sequence after transformation:
X (k) D=(x1(k) d, x2(k) d ..., xm(k)d)。
(2) the smaller the better type of index
The type index for example can be product cost, human cost etc..
Standardized method are as follows:
Former index is smaller, and new index is bigger, the new sequence generated after transformation:
X (k) D=(x1(k) d, x2(k) d ..., xm(k)d)。
(3) index is closer to the better type of some value
Standardized method are as follows:
For former index closer to λ, new index is bigger, and former index distance lambda is remoter, and new index is smaller, raw after transformation
At new sequence: X (k) D=(x1(k) d, x2(k) d ..., xm(k) d), λ is predetermined value.
Step 102 carries out clustering processing to pretreated each index, obtains multiple index clusters.
After carrying out above-mentioned pretreatment to the index in index system, continue to carry out at cluster pretreated index
Reason.
With reference to Fig. 2, this step 102 can specifically be realized by treatment process below and be clustered to pretreated index
Processing:
Step 201, the multiple groups based on index each in index system observe data, calculate in pretreated each index
Multiple grey relational grade numerical value between every two index.
Assuming that an index system includes n index, each index includes m observation data (for example, closing to Business Process System
Key link is analyzed in key process 2005 to 2015, wherein crucial joint includes the index of three links, each finger
Mark includes 10 years data, and granularity is the moon, that is, includes 120 observations.If wherein there is shortage of data, as Length discrepancy refers to
Mark can be handled with the methods of interpolation, sliding window), then have index series as follows:
X1=(x1(1), x1(2) ..., x1(n))
X2=(x2(1), x2(2) ..., x2(n))
Xm=(xm(1), xm(2) ..., xm(n))
To all index i, j ∈ [1, n], X is calculatediWith XjGrey relational grade γ (Xi, Xj), it is abbreviated as γij, can obtain
The degree of association between n index constitutes n rank index degree of association square matrix Γ:
Step 202, according to multiple grey relational grade numerical value between every two index, calculate the average pass between every two index
Connection degree.
Take γijWith γjiMean value as sequence XiWith XjThe average degree of association, be denoted asIt can then obtain
To following average degree of association matrix D:
Step 203, based on the average degree of association between every two index, using predetermined clusters method to pretreated each
Index carries out clustering processing, obtains multiple index clusters, wherein the average degree of association between class between any two index is higher than in class
The average degree of association between any two index.
If G is class, dijFor class GiAnd GjBetween the degree of association, calculation formula are as follows:
Here, most relevance degree is used to calculate the degree of association between class as criterion function, when the application is embodied,
Also the averagely criterion function as cluster such as the degree of association, center of gravity degree of association can be used.
Later, clustering processing is carried out to each index in index system using systemic clustering, steps are as follows:
1) Raw performance has n, and each index constitutes a class by itself, and establishes n class: G0={ G1, G2..., Gn, it is calculated
The mutual average degree of association of n class obtains n rank and is averaged degree of association square matrix D0=dij(n×n).D at this timeij=rij；
2) assume that back has obtained average incidence matrix Dk, k is the number of agglomerative clustering.Find DkIn it is maximum non-
Diagonal entry dij(i ≠ j), then by corresponding GiClass and GjClass merges into a class Gr, thus establish new classification Gk+1=
{G1, G2..., Gr... }；
3) G is recalculatedk+1Average incidence matrix Dk+1；
4) return step 2), it computes repeatedly, until stopping when meeting scheduled stop condition.
Wherein, the stop condition can be with are as follows: (1) all indexs all gather corresponding class；(2) reach required cluster
Number (need to pre-define)；(3) degree of association threshold value, D are setk+1In be not above the element of threshold value.
Step 103, be based on predetermined evaluation method, from each index cluster in select representative degree it is maximum at least one refer to
It is denoted as the representative index in each index cluster.
After above clustering processing, clusters and handled in class continuing with each index, to realize from each finger
At least one maximum index of representative degree is selected in mark cluster as the representative index in each index cluster.
Specifically, with reference to Fig. 3, this step can realize the screening of representative index in class by following treatment process:
301, the observation data based on index calculate the grey relational grade in each index cluster between every two index.
After the clustering processing of previous step, it is assumed that indexs all in index system are poly- for n class G={ G1,
G2..., Gn, successively to class Gi(i=1,2 ..., n) in all index Xj(j=1 2 ..., k) recalculates it two-by-two
Grey relational grade obtains degree of association matrix Γi:
Wherein, k indicates class GiIn include index index quantity.
302, based on the grey relational grade between every two index in each index cluster, each finger in index cluster is calculated
The average degree of association of grey relational grade between mark and other interior each indexs of class, as representative degree of the index in class.
For GiIn index Xj, corresponding association angle value is ΓiIn jth row (γj1, γj2..., γjk), it calculates
The average value of this journey, as index XjAverage correlation degree namely representative degree in class:
303, at least one maximum index of representative degree in each index cluster is selected to cluster as each index
Representative index.
To class GiThe representative degree γ of index is calculated in interior all indexsj(j=1,2 ..., k), and filter out in class
Representative degree γjMaximum one or more (depending on the circumstances) index is as class GiRepresentative index.Other classes then may be used
The representative index in class is filtered out using same method.
Representative index in step 104, each index cluster of merging, the index system after being optimized.
After filtering out at least one of each index cluster representativeness by above step, combinable each index
Representative index in cluster, the index system after finally obtaining optimization.
Based on above scheme it is found that Optimization Method of Index System provided by the present application is analyzed using Grey Correlation Analysis Theory
The index system middle finger target degree of association, and gathered with the method for cluster by each index of the degree of association to index weight
Class, the Cluster-Fusion combined influence factor of various aspects, and according to the average degree of association as index selection mark representative in class
Standard has carried out index screening in class, optimizes index system with this, with higher comprehensive.
With reference to Fig. 4, it is a kind of structural schematic diagram for index system optimization device that another embodiment of the application provides, such as schemes
Shown in 4, which includes:
Pretreatment unit 401, for being pre-processed to each index in index system, so that dimension in index system
Different each indices non-dimensions；
Cluster cell 402 obtains multiple index clusters for carrying out clustering processing to pretreated each index；
Screening unit 403, for be based on predetermined evaluation method, from each index cluster in select representative degree it is maximum to
A few index is as the representative index in each index cluster；
Combining unit 404, the index system for merging the representative index in each index cluster, after being optimized.
In an embodiment of the embodiment of the present application, the pretreatment unit 401 is specifically used for: being based on pre-defined rule pair
Each index in index system is grouped, and obtains multiple index groups；Scheduled nothing is carried out to the index in each index group
Dimensionization processing.
In an embodiment of the embodiment of the present application, the pretreatment unit 401 is based on pre-defined rule to index system
In each index be grouped, obtain multiple index groups, specifically include: by the value type of index to each in index system
A index is grouped, and obtains multiple index groups；The value type of the index includes for indicating what index value was the bigger the better
The first kind, for indicating the smaller the better Second Type of index value, for indicating that index value is better closer to a certain predetermined value
Third type.
In an embodiment of the embodiment of the present application, the cluster cell 402 is specifically used for:
Multiple groups based on index each in index system observe data, calculate every two in pretreated each index and refer to
Multiple grey relational grade numerical value between mark；According to multiple grey relational grade numerical value between every two index, every two index is calculated
Between the average degree of association；Based on the average degree of association between every two index, using predetermined clusters method to pretreated each
Index carries out clustering processing, obtains multiple index clusters, wherein the average degree of association between class between any two index is higher than in class
The average degree of association between any two index.
In an embodiment of the embodiment of the present application, the screening unit 403 is specifically used for:
Observation data based on index calculate the grey relational grade in each index cluster between every two index；Based on every
Grey relational grade in a index cluster between every two index calculates each index and other interior each indexs of class in index cluster
Between grey relational grade the average degree of association, as representative degree of the index in class；Select in each index cluster representative degree most
Representative index of at least one the big index as each index cluster.
For index system optimization device disclosed by the embodiments of the present invention, since it refers to disclosed in above embodiments
Mark system optimization method is corresponding, so being described relatively simple, related similarity refers to above embodiments middle finger standard type
It is the explanation of optimization method part, and will not be described here in detail.
In conclusion the Optimization Method of Index System and device of the application, clustering is mutually tied with grey correlation analysis
It closes, the measurement first with grey relational grade as similitude between index, and using hierarchial-cluster analysis to existing index system
It is clustered, so that index to be divided into different major class, every one kind index all describes the different aspect of system.In every one kind
In further recalculate grey relational grade between index, take the maximum one or more representatives as such of the average degree of association
Property represent, finally merge the index system after all kinds of representative indexs is optimized, compared with the prior art, the finger of the application
Mark system optimization method and device, at least has following technical effect that
(1) the application is in building index optimization system, for the first time using grey relational grade as the degree of similitude between index
Amount has novelty；
(2) model simple that the application is proposed is illustrated, and is easily mastered, versatile；
(3) the index optimization scheme of the application building is by the Cluster-Fusion combined influence factor of various aspects, and from poly-
Representative index is selected in each class that class generates, is had comprehensive.
It should be noted that all the embodiments in this specification are described in a progressive manner, each embodiment weight
Point explanation is the difference from other embodiments, and the same or similar parts between the embodiments can be referred to each other.
For convenience of description, it describes to be divided into various modules when system above or device with function or unit describes respectively.
Certainly, the function of each unit can be realized in the same or multiple software and or hardware in carrying out the present invention.
As seen through the above description of the embodiments, those skilled in the art can be understood that the present invention can
It realizes by means of software and necessary general hardware platform.Based on this understanding, technical solution of the present invention essence
On in other words the part that contributes to existing technology can be embodied in the form of software products, the computer software product
It can store in storage medium, such as ROM/RAM, magnetic disk, CD, including some instructions are used so that a computer equipment
(can be personal computer, server or the network equipment etc.) executes the certain of each embodiment or embodiment of the invention
Method described in part.
Finally, it is to be noted that, herein, such as first, second, third and fourth or the like relational terms
It is only used to distinguish one entity or operation from another entity or operation, without necessarily requiring or implying these
There are any actual relationship or orders between entity or operation.Moreover, the terms "include", "comprise" or its is any
Other variants are intended to non-exclusive inclusion, so that including the process, method, article or equipment of a series of elements
Include not only those elements, but also including other elements that are not explicitly listed, or further includes for this process, side
Method, article or the intrinsic element of equipment.In the absence of more restrictions, limited by sentence "including a ..."
Element, it is not excluded that there is also other identical elements in the process, method, article or apparatus that includes the element.
The above is only a preferred embodiment of the present invention, it is noted that for the ordinary skill people of the art
For member, various improvements and modifications may be made without departing from the principle of the present invention, these improvements and modifications are also answered
It is considered as protection scope of the present invention.
Claims (10)
1. a kind of Optimization Method of Index System characterized by comprising
Each index in index system is pre-processed, so that the different each index of dimension is immeasurable in index system
Guiding principle；
Clustering processing is carried out to pretreated each index, obtains multiple index clusters；
Based on predetermined evaluation method, at least one maximum index of representative degree is selected from each index cluster as described every
Representative index in a index cluster；
Merge the representative index in each index cluster, the index system after being optimized.
2. the method according to claim 1, wherein each index in index system is located in advance
Reason, so that the different each indices non-dimension of dimension in index system, comprising:
Each index in index system is grouped based on pre-defined rule, obtains multiple index groups；
Scheduled nondimensionalization processing is carried out to the index in each index group.
3. according to the method described in claim 2, it is characterized in that, it is described based on pre-defined rule to each finger in index system
Mark is grouped, and obtains multiple index groups, comprising:
Each index in index system is grouped by the value type of index, obtains multiple index groups；
The value type of the index includes for indicating the first kind that index value is the bigger the better, for indicating that index value is smaller
Better Second Type, for indicating the index value third type better closer to a certain predetermined value.
4. the method according to claim 1, wherein described carry out at cluster pretreated each index
Reason obtains multiple index clusters, comprising:
Multiple groups based on index each in index system observe data, calculate in pretreated each index between every two index
Multiple grey relational grade numerical value；
According to multiple grey relational grade numerical value between every two index, the average degree of association between every two index is calculated；
Based on the average degree of association between every two index, pretreated each index is clustered using predetermined clusters method
Processing obtains multiple index clusters, wherein the average degree of association between class between any two index is higher than any two index in class
Between the average degree of association.
5. being clustered the method according to claim 1, wherein described be based on predetermined evaluation method from each index
In select at least one representative maximum index as the representative index in each index cluster, comprising:
Observation data based on index calculate the grey relational grade in each index cluster between every two index；
Based on the grey relational grade between every two index in each index cluster, calculate in index cluster in each index and class
The average degree of association of grey relational grade between other each indexs, as representative degree of the index in class；
Select representativeness of at least one the maximum index of representative degree as each index cluster in each index cluster
Index.
6. a kind of index system optimization device characterized by comprising
Pretreatment unit, for being pre-processed to each index in index system, so that dimension is different in index system
Each indices non-dimension；
Cluster cell obtains multiple index clusters for carrying out clustering processing to pretreated each index；
Screening unit, for be based on predetermined evaluation method, from each index cluster in select representative degree it is maximum at least one
Index is as the representative index in each index cluster；
Combining unit, the index system for merging the representative index in each index cluster, after being optimized.
7. device according to claim 6, which is characterized in that the pretreatment unit is specifically used for:
Each index in index system is grouped based on pre-defined rule, obtains multiple index groups；
Scheduled nondimensionalization processing is carried out to the index in each index group.
8. device according to claim 7, which is characterized in that the pretreatment unit is based on pre-defined rule to index system
In each index be grouped, obtain multiple index groups, specifically include:
Each index in index system is grouped by the value type of index, obtains multiple index groups；
The value type of the index includes for indicating the first kind that index value is the bigger the better, for indicating that index value is smaller
Better Second Type, for indicating the index value third type better closer to a certain predetermined value.
9. device according to claim 6, which is characterized in that the cluster cell is specifically used for:
Multiple groups based on index each in index system observe data, calculate in pretreated each index between every two index
Multiple grey relational grade numerical value；
According to multiple grey relational grade numerical value between every two index, the average degree of association between every two index is calculated；
Based on the average degree of association between every two index, pretreated each index is clustered using predetermined clusters method
Processing obtains multiple index clusters, wherein the average degree of association between class between any two index is higher than any two index in class
Between the average degree of association.
10. device according to claim 6, which is characterized in that the screening unit is specifically used for:
Observation data based on index calculate the grey relational grade in each index cluster between every two index；
Based on the grey relational grade between every two index in each index cluster, calculate in index cluster in each index and class
The average degree of association of grey relational grade between other each indexs, as representative degree of the index in class；
Select representativeness of at least one the maximum index of representative degree as each index cluster in each index cluster
Index.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811528291.6A CN109376966A (en) | 2018-12-13 | 2018-12-13 | A kind of Optimization Method of Index System and device |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811528291.6A CN109376966A (en) | 2018-12-13 | 2018-12-13 | A kind of Optimization Method of Index System and device |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN109376966A true CN109376966A (en) | 2019-02-22 |
Family
ID=65373756
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201811528291.6A Pending CN109376966A (en) | 2018-12-13 | 2018-12-13 | A kind of Optimization Method of Index System and device |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109376966A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111985815A (en) * | 2020-08-21 | 2020-11-24 | 国网能源研究院有限公司 | Method and device for screening energy and power operation evaluation indexes |
-
2018
- 2018-12-13 CN CN201811528291.6A patent/CN109376966A/en active Pending
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111985815A (en) * | 2020-08-21 | 2020-11-24 | 国网能源研究院有限公司 | Method and device for screening energy and power operation evaluation indexes |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN107256426B (en) | Medicine safety confidence index system based on intelligent algorithm and construction method | |
| CN106022477A (en) | Intelligent analysis decision system and method | |
| Calders et al. | What is data mining and how does it work? | |
| CN110675243A (en) | Machine learning-fused credit prediction overdue method and system | |
| Piad et al. | Predicting IT employability using data mining techniques | |
| CN110490438B (en) | Simplified strategy recommendation method for digital twinning capability of industrial assembly line workshop | |
| CN112419029B (en) | Similar financial institution risk monitoring method, risk simulation system and storage medium | |
| Li et al. | Application of improved AHP-BP neural network in CSR performance evaluation model | |
| CN110765268A (en) | Client appeal-based accurate distribution network investment strategy method | |
| Ghazal et al. | Data Mining and Exploration: A Comparison Study among Data Mining Techniques on Iris Data Set | |
| Gianazza et al. | Evaluation of air traffic complexity metrics using neural networks and sector status | |
| Rabbi et al. | An Approximation For Monitoring The Efficiency Of Cooperative Across Diverse Network Aspects | |
| Lee et al. | Foresight of promising technologies for healthcare-IoT convergence service by patent analysis | |
| CN111738843A (en) | Quantitative risk evaluation system and method using running water data | |
| Silva | Metrics for evaluating performance in document analysis: application to tables | |
| CN109376966A (en) | A kind of Optimization Method of Index System and device | |
| CN115660221B (en) | Oil and gas reservoir economic recoverable reserve assessment method and system based on hybrid neural network | |
| CN109146232A (en) | Stock market's investor sentiment of principal component and the coefficient of variation acts on behalf of index selection method | |
| CN106778252A (en) | Intrusion detection method based on rough set theory Yu WAODE algorithms | |
| CN113920366A (en) | Comprehensive weighted main data identification method based on machine learning | |
| CN109934489B (en) | Power equipment state evaluation method | |
| CN114912772A (en) | Urban right transparency differential evaluation system matching method and system based on urban economic classification analysis | |
| CN113705679A (en) | Student score prediction method based on hypergraph neural network | |
| CN112506930A (en) | Data insight platform based on machine learning technology | |
| Hariharan et al. | An interpretable predictive model of vaccine utilization for Tanzania |
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 | ||
| TA01 | Transfer of patent application right |
Effective date of registration: 20190709 Address after: 100033 West Chang'an Avenue, Xicheng District, Xicheng District, Beijing Applicant after: State Grid Corporation of China Applicant after: STATE GRID INFORMATION & TELECOMMUNICATION GROUP CO., LTD. Address before: 100192 room 710, research building 15, Xiaoying East Road, Qinghe, Haidian District, Beijing Applicant before: Beijing China Power Information Technology Co., Ltd. Applicant before: STATE GRID INFORMATION & TELECOMMUNICATION GROUP CO., LTD. Applicant before: State Grid Corporation of China |
|
| TA01 | Transfer of patent application right | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20190222 |
|
| RJ01 | Rejection of invention patent application after publication |