Research On Methods For Multi-attribute Group Decision Making With Interval Numbers | Posted on:2015-12-01 | Degree:Master | Type:Thesis | Country:China | Candidate:X L Xie | Full Text:PDF | GTID:2309330431985381 | Subject:Management Science and Engineering | Abstract/Summary: | PDF Full Text Request | Multi-attribute group decision making is a cross product of multi-attributedecision-making and group decision making, which is not only very common in socialpractice but also with a wide range of applications. As one of the branches of scientificmanagement decisions, research on theory and method of multi-attribute group decisionmaking is causing growing concern in various fields. The idea is in a decision problem,different decision makers give assessment information for each attribute of each program thatconstitutes a group decision making matrix. Then certain methods are proposed to gatherinformation and a finite number of programs are sorted, selecting the optimal decision results.Large data sets are in the information age today, due to the complexity of the issue,decision making systems continue to grow and subjective consciousness of the decisionmakers, decision makers often hesitate during the evaluation process. Therefore, it is moreappropriate to describe the property in the form of interval numbers. This paper studiesmulti-attribute group decision making that attribute decision information given in the form ofinterval numbers and attribute weights of multi-attribute known or unknown. That is to say, itpays attention to gathering interval numbers decision-making information of attributes ofprograms evaluated by many decision makers, determining attribute weights and sorting theprograms. This paper solves the issue of research scientifically and effectively based onlogistics optimization theory and artificial intelligence algorithms.Gather interval numbers decision-making information based on plant growth simulationalgorithm. The paper introduces definitions and operational rules for interval numbers andproposes the concept and property of clutch degree. Then it is introduced into the concept ofgroup decision making, given the related symbols. After elaborating simulation process anditerative steps of plant growth simulation algorithm, the algorithm is improved to determinethe initial growing point by Sierpinski carpet. The original model which is used to gatherinterval numbers called Steiner problem is presented, while applying its ideas and methods ofcomputation to gather information on group decision making. This paper proposes thenonlinear programming model to gather information based on plant growth simulationalgorithm. To our knowledge, the application of plant growth simulation algorithm in the fieldof management decisions to collect program on each attribute group decision makingpreference interval number is the first trial in literature. The results further prove that thisapproach has significant feasibility and effectiveness. Apply projection theory and improved TOPSIS method for multi-attribute group decisionmaking with interval numbers. The paper discusses the basic theory of interval numberprojection. Moreover, indicators are divided into efficiency attributes and cost attributes inorder to deal with the original interval numbers. The definitions of interval positive andnegative ideal point are given and then the decision making processes of projection theory areput forward to sort programs. However, the projection theory uses only positive ideal point tointerval number and it is therefore necessary to strengthen the robustness of the results of theevaluation, the paper presents the TOPSIS method as well as taking into account virtualinterval number to improve it. In all, the research on multi-attribute group decision makingbased on improved TOPSIS method is proposed to make the way of judging the pros and consof the programs more reliable.Determine the attribute weights based on entropy weight method and research onmulti-attribute group decision making with interval numbers based on possibility degree.Renewal of group decision making preferences matrix by plant growth simulation algorithmfor information gathering, there is entropy weight in Evolution of Decision Sciences fromentropy theory. According to the degree of change within the overall relative differences in thevariability, the paper puts forward objective attribute weights determined by entropy weightmethod. Then it gives the definitions of possibility degree and possibility degree matrix forcomparing interval numbers and both analyzes and proves the nature of possible degreeformula. When obtaining the possibility degree preferences matrix, the program can be sortedby priority weight vector method. Therefore, this paper study the ways and the steps onmulti-attribute group decision making with interval numbers which its attribute weights arecompletely unknown on account of the entropy weight method combined with possibilitydegree and priority weight vector method.Study on interval numbers group decision-making of venture capital in technologyenterprises. The paper creates the evaluation index system of venture capital in technologyenterprises and empirical analysis on multi-attribute group decision making under intervalnumbers information. The results show that the methods of this paper are not only same asother literatures’ final sorting with different calculation methods, but also more convenientand flexible, which are meaningful for practical use on venture capital. | Keywords/Search Tags: | interval number, multi-attribute group decision making, plant growthsimulation algorithm, projection theory, improved TOPSIS, entropy weight method, possibledegree | PDF Full Text Request | Related items |
| |
|