CN106611239A - Improved K-MEANS algorithm for solving material purchasing problem of manufacturing industry - Google Patents
Improved K-MEANS algorithm for solving material purchasing problem of manufacturing industry Download PDFInfo
- Publication number
- CN106611239A CN106611239A CN201610835193.1A CN201610835193A CN106611239A CN 106611239 A CN106611239 A CN 106611239A CN 201610835193 A CN201610835193 A CN 201610835193A CN 106611239 A CN106611239 A CN 106611239A
- Authority
- CN
- China
- Prior art keywords
- cluster
- supplier
- cost
- algorithm
- improved
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q10/00—Administration; Management
- G06Q10/04—Forecasting or optimisation specially adapted for administrative or management purposes, e.g. linear programming or "cutting stock problem"
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/23—Clustering techniques
- G06F18/232—Non-hierarchical techniques
- G06F18/2321—Non-hierarchical techniques using statistics or function optimisation, e.g. modelling of probability density functions
- G06F18/23213—Non-hierarchical techniques using statistics or function optimisation, e.g. modelling of probability density functions with fixed number of clusters, e.g. K-means clustering
-
- 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
Abstract
An improved K-MEANS algorithm for solving a material purchasing problem of manufacturing industry comprises the steps of clustering a supplier data set; selecting optimal cluster to perform cost consideration; calculating cost of a supplier; calculating an appropriate index of a merchant; calculating an optimal cost merchant; and finally, outputting a result. By the method, the data set is clustered, attribute similarity consideration is performed on cluster optimal set, thus, the calculation quantity of the algorithm is reduced, the search efficiency of the algorithm is improved, and on the other hand, the practicability of algorithm is improved; by taking the kind and the number of standard purchasing materials as cluster number, the algorithm is simple and effective, and the complexity of the algorithm is reduced; the distances between suppliers and the cluster centers in two supplier sets are depicted by material kinds and numbers to express dissimilarity degree, and the algorithm is easy to understand and is high in practicability; and cost consideration is performed by selecting optimal cluster, the sensitivity of the algorithm on an abnormal value is reduced, meanwhile, the range of seeking optimal solution by the algorithm is reduced, and the solution obtained through seeking in the optimal class will be more approximate to an ideal solution.
Description
Technical field
The present invention relates to business administration field, more particularly to algorithm manufacturing industry material procurement is solved the problems, such as.
Background technology
With world market integration and the arrival of information age, the proportion of enterprise procurement is greatly increased.In global model
In enclosing, during the product of industrial undertaking is constituted, the raw material and parts cost average level of buying is more than 60%.From world's model
For enclosing, for a typical enterprise, purchase cost (including raw material, parts) will account for 60%.And in the industry of China
Enterprise, the purchase cost of various goods and materials will account for the 70% of enterprise marketing cost.Obviously purchase cost is the master in business administration
Body and core, buying is the part of " most worthy " in business administration.In addition, issued within 1999 according to State Economic and Trade Commission
Relevant data, if large and medium-sized state-owned enterprise reduces every year purchase cost 2%-3%, can increase the remote RMB of benefit more than 500 hundred million,
Equivalent to the summation that people enterprise of state-owned industry in 1997 generates profit.Therefore, buying increasingly receives the concern and attention of people,
Research to buying is also referred to as one of hot issue of today's society.
However, a kind of efficient intelligent optimization method of research is very meaningful to solve the problems, such as material procurement.Enterprise procurement
Main target be reduce entreprise cost.
K_means algorithms have a very wide range of applications in mass data process, and its main thought is by iteration
Process is divided into different classifications data set so that the criterion function for evaluating clustering performance is optimal, so that generate
It is compact in each cluster, it is independent between class.K_means algorithms have the advantages that the calculating time is short, speed is fast, easily explain.But it is right
Exceptional value is sensitive, not accurate enough to the solution of some problems.
The content of the invention
For the above-mentioned deficiency of prior art, the technical problem to be solved in the present invention is to provide a kind of improved K_MEANS
Algorithm is to solving the problems, such as manufacturing industry material procurement.
The purpose of the present invention is to overcome problems of the prior art:For manufacturing industry purchasing problem, K_means algorithms
Computational accuracy is not high enough, sensitive to exceptional value, and manufacturing industry material procurement problem has that many compromises are not solved.
The technical scheme that adopted for achieving the above object of the present invention is:A kind of improved K_MEANS algorithms are to solving system
Make industry material procurement problem.
The step of algorithm, is as follows:
Step 1:Initialization data set.Initialization purchaser data set P, attribute (material) data set A, purchase material standard
X。
Step 2:Supplier data collection is clustered.With improved K_Means algorithms to supplier data clustering,
Gather for a class with the higher supplier of similarity, be specifically:
(1) number of clusters k is determined.
(2) cluster centre is selected.It is random that k supplier is selected in supply quotient set as initial cluster center.
(3) distinctiveness ratio is calculated.Two are portrayed with material variety quantity with supplier and cluster centre in supply quotient set
Distance representing distinctiveness ratio.
(4) cluster.Each supplier is clustered in the cluster centre minimum with its distinctiveness ratio.
(5) cluster meansigma methodss are calculated.The material variety meansigma methodss of all suppliers in each cluster are calculated, and this is put down
Average is used as new cluster centre.
(6) (3), (4) are performed repeatedly, until cluster centre is no longer moved on a large scale or clusters number of times requirement is reached
Till.
(7) output cluster.
Step 3:Optimum cluster is selected to carry out cost viewpoint.The selection rule of optimum cluster is as follows:
(1) all class centers are calculated to the distance of standard purchase price amount.
(2) the minimum cluster of chosen distance counts the Supplier Number that optimum cluster is included as optimum cluster
Nbest。
(3) cost viewpoint is as described in step 4.
Step 4:Calculate the cost of each supplier of optimum cluster the inside.
Step 5:Calculate the appropriate index of businessman.
Step 6:Select the maximum material supplier of fitness.
Step 7:Optimal cost businessman is found out in calculating.
Step 8:Algorithm terminates, and exports optimum.
The invention has the beneficial effects as follows:
1st, data set is clustered first, then collection optimum to cluster carries out attributes similarity and considers, less algorithm
Amount of calculation, improve algorithm search efficiency, on the other hand, improve the practical degree of algorithm solution.
2nd, material variety number is bought as number of clusters using standard, it is simple effective, reduce the complexity of algorithm.
3rd, portray two distances with supplier and cluster centre in supply quotient set with material variety quantity to represent
Distinctiveness ratio, it is easy to understand, practicality is high.
4th, select optimum cluster to carry out cost viewpoint, sensitivity of the algorithm to exceptional value is reduced, while reducing calculation
Method searches the scope of optimal solution, while searching in premium class, the solution for obtaining can be closer with ideal solution.
5th, by the cost of calculating optimum cluster each supplier of the inside, compare, obtain optimal solution.The algorithm solution of raising
Degree of accuracy.
6th, by the appropriate index of businessman, using into the appropriate index for portraying supplier originally, make last solution relatively reliable.
Description of the drawings
Fig. 1 is a kind of improved K_MEANS algorithms to solving the problems, such as manufacturing industry material procurement.
Fig. 2 is supplier's clustering algorithm figure.
Specific embodiment
In order that the objects, technical solutions and advantages of the present invention become more apparent, carry out below in conjunction with algorithm flow chart
In detail, illustrate.
First, the description of material procurement problem
Initialization purchaser data set P, attribute (material) data set A, purchase material standard x.It is provided with K class goods and materials X={ Xi
| i=1,2 ..., K }.Provide P={ P from supplier respectivelyj| i=1,2 ..., N } place's buying, every kind of goods and materials have buying index,
Such as price and freight.When being purchased, every kind of goods and materials all select an index as leading indicator, and other indexs are made
For reference index.The supplier of different goods and materials how is selected, a procurement scheme X={ X is formulated1, X2..., Xk, adopt in satisfaction
On the premise of the constraintss such as purchase amount, ordering period and fund, make total purchase cost minimum.Set up plan model as follows:
Object function:
Wherein qi=f (x1, x2..., xm) be the i-th class goods and materials amount of purchase, qixijRepresent the i-th class goods and materials with regard to index j
Cost, xijRepresent j-th buying desired value, C (Xi) represent the i-th class goods and materials purchase cost, C (x*) it is optimum procurement scheme
x*Cost, i.e. optimal cost.
Constraints:
1) amount of purchase constraint:Buyer and supplier allow each index the interval of a floating, i.e.,:xij∈
[T1(j), T2(j)], i=1,2 ..., n;J=1,2 ..., m
T1(j), T2J () is respectively the lower limit and the upper limit of index.
2) limited fund:Need to meet per the procurement payment of class goods and materials:
qi=f (x1, x2..., xm) it is k homogeneous equation, cost function c (qi) reflect purchase cost and amount of purchase it
Between restriction relation,L is purchase cost.
2nd, specific implementation step
Step 1:Initialization data set.Initialization purchaser data set P, attribute (material) data set A, purchase material standard
X.P={ Pj| i=1,2 ..., N }, X={ Xi| i=1,2 ..., K }, A={ Ai| i=1,2 ..., NK }
Step 2:Supplier data collection is clustered.With improved K_Means algorithms to supplier data clustering,
Gather for a class with the higher supplier of similarity, be specifically:
(1) number of clusters k is determined.
(2) cluster centre is selected.It is random that k supplier is selected in supply quotient set as initial cluster center.It is designated as V
={ vc| c=1,2 ... k }.
(3) distinctiveness ratio is calculated.Two are portrayed with material variety quantity with supplier and cluster centre in supply quotient set
Distance representing distinctiveness ratio.Specifically it is calculated as:
Wherein, niFor the material variety number that supplier i can be provided, nicCentered on put the things that can be provided of supplier c
Material species number.
(4) cluster.Each supplier is clustered in the cluster centre minimum with its distinctiveness ratio.
(5) cluster meansigma methodss are calculated.The material variety meansigma methodss of all suppliers in each cluster are calculated, and this is put down
Average is used as new cluster centre.
(6) (3), (4) are performed repeatedly, until cluster centre is no longer moved on a large scale or clusters number of times requirement is reached
Till.
(7) output cluster.
Step 3:Optimum cluster is selected to carry out cost viewpoint.The selection rule of optimum cluster is as follows:
(1) all class centers are calculated to the distance of standard purchase price amount.Computing formula is:
Dc=| (nic-k)|
(2) the minimum cluster of chosen distance counts the Supplier Number that optimum cluster is included as optimum cluster
Nbest。
(3) cost viewpoint is as described in step 4.
Step 4:Calculate the cost of each supplier of optimum cluster the inside.
Step 5:Calculate the appropriate index of businessman.
Step 6:Select the maximum material supplier of fitness.Selection criterion is as follows:
The corresponding suppliers of wherein Better are designated as p.
Step 7:Optimal cost businessman is found out in calculating.Computational methods are as follows:
Object function:
Wherein qi=f (x1, x2..., xk) be the i-th class goods and materials amount of purchase, qixijRepresent the i-th class goods and materials with regard to index j
Cost, xijRepresent j-th buying desired value, C (Xi) represent the i-th class goods and materials purchase cost, C (x*) it is optimum procurement scheme
x*Cost, i.e. optimal cost.
Constraints:
1) amount of purchase constraint:Buyer and supplier allow each index the interval of a floating, i.e.,:xij∈
[T1(j), T2(j)], i=1,2 ..., k;J=1,2 ..., Nbest
T1(j), T2J () is respectively the lower limit and the upper limit of index.
2) limited fund:Need to meet per the procurement payment of class goods and materials:
qi=f (x1, x2..., xk) it is k homogeneous equation, cost function c (qi) reflect purchase cost and amount of purchase it
Between restriction relation,L is purchase cost.
Wherein optimal cost supplier is designated as q.
Step 8:Algorithm terminates, and exports optimum.
Claims (7)
1. a kind of improved K_MEANS algorithms are to solving the problems, such as manufacturing industry material procurement, the present invention relates to business administration field, tool
It is related to solve the problems, such as manufacturing industry material procurement with algorithm body, it is characterized in that, comprises the steps:
Step 1:Initialization data set, initializes purchaser data set P, attribute (material) data set A, purchases material standard x
Step 2:Supplier data collection is clustered, with improved K_Means algorithms to supplier data clustering, having
The higher supplier of similarity gathers for a class, is specifically:
(1)Determine number of clusters k
(2)Cluster centre is selected, k supplier is selected as initial cluster center in supply quotient set at random
(3)Calculate distinctiveness ratio, portrayed with material variety quantity two with supply quotient set supplier and cluster centre away from
From representing distinctiveness ratio
(4)Cluster, each supplier is clustered in the cluster centre minimum with its distinctiveness ratio
(5)Cluster meansigma methodss are calculated, the material variety meansigma methodss of all suppliers in each cluster are calculated, and by this meansigma methods
As new cluster centre
(6)Perform repeatedly(3)、(4), until cluster centre is no longer moved on a large scale or clustered till number of times reaches requirement
(7)Output cluster
Step 3:Optimum cluster is selected to carry out cost viewpoint, the selection rule of optimum cluster is as follows:
(1)All class centers are calculated to the distance of standard purchase price amount
(2)The minimum cluster of chosen distance counts the Supplier Number that optimum cluster is included as optimum cluster
(3)Cost viewpoint is as described in step 4
Step 4:Calculate the cost of each supplier of optimum cluster the inside
Step 5:Calculate the appropriate index of businessman
Step 6:Select the maximum material supplier of fitness
Step 7:Optimal cost businessman is found out in calculating
Step 8:Algorithm terminates, and exports optimum.
2., according to a kind of improved K_MEANS algorithms described in claim 1 to solving the problems, such as manufacturing industry material procurement, it is special
Levying is, the specific implementation step in the above step 2 is as follows:
Step 2:Supplier data collection is clustered, with improved K_Means algorithms to supplier data clustering, having
The higher supplier of similarity gathers for a class, is specifically:
(1)Determine number of clusters k
(2)Cluster centre is selected, selects k supplier as initial cluster center in supply quotient set at random, be designated as
(3)Calculate distinctiveness ratio, portrayed with material variety quantity two with supply quotient set supplier and cluster centre away from
From representing distinctiveness ratio, specifically it is calculated as:
Wherein,For the material variety number that supplier i can be provided,Centered on put the materials that can be provided of supplier c
Species number
(4)Cluster, each supplier is clustered in the cluster centre minimum with its distinctiveness ratio
(5)Cluster meansigma methodss are calculated, the material variety meansigma methodss of all suppliers in each cluster are calculated, and by this meansigma methods
As new cluster centre
(6)Perform repeatedly(3)、(4), until cluster centre is no longer moved on a large scale or clustered till number of times reaches requirement
(7)Output cluster.
3., according to a kind of improved K_MEANS algorithms described in claim 1 to solving the problems, such as manufacturing industry material procurement, it is special
Levying is, the specific implementation process in the above step 3 is as follows:
Step 3:Optimum cluster is selected to carry out cost viewpoint, the selection rule of optimum cluster is as follows:
(1)All class centers are calculated to the distance of standard purchase price amount, computing formula is:
(2)The minimum cluster of chosen distance counts the Supplier Number that optimum cluster is included as optimum cluster
(3)Cost viewpoint.
4. according to a kind of improved K_MEANS algorithms described in claim 3 to solving the problems, such as manufacturing industry material procurement, wherein
Cost viewpoint is characterized in that:
Calculate the cost of each supplier of optimum cluster the inside
。
5., according to a kind of improved K_MEANS algorithms described in claim 1 to solving the problems, such as manufacturing industry material procurement, it is special
Levying is, the specific implementation process in the above step 5 is as follows:
Step 5:Calculate the appropriate index of businessman
。
6., according to a kind of improved K_MEANS algorithms described in claim 1 to solving the problems, such as manufacturing industry material procurement, it is special
Levying is, the specific implementation process in the above step 6 is as follows:
Step 6:The maximum material supplier of fitness is selected, selection criterion is as follows:
The corresponding suppliers of wherein Better are designated as p.
7., according to a kind of improved K_MEANS algorithms described in claim 1 to solving the problems, such as manufacturing industry material procurement, it is special
Levying is, the specific implementation process in the above step 7 is as follows:
Step 7:Optimal cost businessman is found out in calculating, and computational methods are as follows:
Object function:
WhereinFor adopting for the i-th class goods and materials
Purchase amount,Cost of the i-th class goods and materials with regard to index j is represented,J-th buying desired value is represented,Represent the i-th class
The purchase cost of goods and materials,For optimum procurement schemeCost, i.e. optimal cost
Constraints:
1)Amount of purchase is constrained:Buyer and supplier allow each index the interval of a floating, i.e.,:
The respectively lower limit and the upper limit of index
2)Limited fund:Need to meet per the procurement payment of class goods and materials:
It is k homogeneous equation, cost functionReflect between purchase cost and amount of purchase
Restriction relation,For purchase cost
Wherein optimal cost supplier is designated as q.
Applications Claiming Priority (2)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201610595239 | 2016-07-26 | ||
CN2016105952397 | 2016-07-26 |
Publications (1)
Publication Number | Publication Date |
---|---|
CN106611239A true CN106611239A (en) | 2017-05-03 |
Family
ID=58615313
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201610835193.1A Pending CN106611239A (en) | 2016-07-26 | 2016-09-20 | Improved K-MEANS algorithm for solving material purchasing problem of manufacturing industry |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN106611239A (en) |
Cited By (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN112035978A (en) * | 2020-09-03 | 2020-12-04 | 哈尔滨理工大学 | Cutter parameter optimization design method and system |
CN114417964A (en) * | 2021-12-10 | 2022-04-29 | 中国卫通集团股份有限公司 | Satellite operator classification method and device and electronic equipment |
US20230052034A1 (en) * | 2021-08-13 | 2023-02-16 | Edgeverve Systems Limited | Method and system for analyzing process flows for a process performed by users |
-
2016
- 2016-09-20 CN CN201610835193.1A patent/CN106611239A/en active Pending
Cited By (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN112035978A (en) * | 2020-09-03 | 2020-12-04 | 哈尔滨理工大学 | Cutter parameter optimization design method and system |
CN112035978B (en) * | 2020-09-03 | 2022-07-12 | 哈尔滨理工大学 | Cutter parameter optimization design method and system |
US20230052034A1 (en) * | 2021-08-13 | 2023-02-16 | Edgeverve Systems Limited | Method and system for analyzing process flows for a process performed by users |
US11847598B2 (en) * | 2021-08-13 | 2023-12-19 | Edgeverve Systems Limited | Method and system for analyzing process flows for a process performed by users |
CN114417964A (en) * | 2021-12-10 | 2022-04-29 | 中国卫通集团股份有限公司 | Satellite operator classification method and device and electronic equipment |
CN114417964B (en) * | 2021-12-10 | 2023-09-12 | 中国卫通集团股份有限公司 | Satellite operator classification method and device and electronic equipment |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
WO2020215751A1 (en) | Recommendation method, electronic device and readable storage medium | |
JP6134444B2 (en) | Method and system for recommending information | |
US7885859B2 (en) | Assigning into one set of categories information that has been assigned to other sets of categories | |
TWI512653B (en) | Information providing method and apparatus, method and apparatus for determining the degree of comprehensive relevance | |
CN102902691B (en) | Recommend method and system | |
CN110782266B (en) | Commodity sales prediction and model training method, device and equipment | |
WO2021139164A1 (en) | Sequential recommendation method based on long-term interest and short-term interest | |
TWI519976B (en) | Category misplaced recognition methods and devices | |
WO2020238229A1 (en) | Transaction feature generation model training method and devices, and transaction feature generation method and devices | |
TW201606537A (en) | Visual interactive search | |
WO2016150354A1 (en) | Method and system for classifying users of e-commerce platform | |
WO2022156529A1 (en) | Commodity recommendation method and apparatus for enterprise user | |
CN102841946A (en) | Commodity data retrieval sequencing and commodity recommendation method and system | |
CN105468628B (en) | A kind of sort method and device | |
CN108734587A (en) | The recommendation method and terminal device of financial product | |
CN106611239A (en) | Improved K-MEANS algorithm for solving material purchasing problem of manufacturing industry | |
WO2019072098A1 (en) | Method and system for identifying core product terms | |
US10832305B1 (en) | System and method for image processing and searching for classification in a product database | |
CN107093122B (en) | Object classification method and device | |
CN115496566A (en) | Regional specialty recommendation method and system based on big data | |
WO2020119017A1 (en) | System and method for achieving data asset sensing and pricing functions in big data background | |
CN106611240A (en) | K_Means algorithm based on evolutionary algorithm for solving material purchasing problem in manufacturing industry | |
CN113763089A (en) | Article recommendation method and device and computer-readable storage medium | |
CN110020918B (en) | Recommendation information generation method and system | |
CN110189188A (en) | Merchandise control method, apparatus, computer equipment and storage medium |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
PB01 | Publication | ||
PB01 | Publication | ||
SE01 | Entry into force of request for substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
WD01 | Invention patent application deemed withdrawn after publication |
Application publication date: 20170503 |
|
WD01 | Invention patent application deemed withdrawn after publication |