Research On Multi-Attribute Group Decision Making Methods With Probability Linguistic And Theirs Applications Based On Consensus Perspective

Multi-Attribute Group Decision Making(MAGDM)method is an important decision analysis technology,which has already been proven to be effective and suitable to be applied in the investment scheme evaluation,supplier selection,industrial site selection,human resource performance evaluation,etc.In practical decision making process,decision makers(DMs)can use linguistic terms such as"good","fair" or "poor" to express their preferences over the considered alternatives,and make the optimal decision by using some proper decision making methods.However,the uncertainty and vagueness of qualitative indicators are difficult to be expressed by using just one linguistic term,the probabilistic linguistic term sets includes several possible linguistic terms and the associated probabilistic information,which is a generalization of linguistic label based on hesitant fuzzy linguistic term sets,can be used to depict the linguistic evaluation for an alternative with respect to a attribute and the hesitation degree on such evaluation simulataneously.Comparing to the traditional linguistic term,the probabilistic linguistic term sets can size the fuzziness and uncertainty of human thinking and is more suitable for solving some complex decision-making problems.At present,with respect to the MAGDM methds with probabilistic linguistic term sets,lacking the systematic theoretical foundation.Consequently,these problems are investigatied systematically in this thesis,the research contents cover the whole process of MAGDM,including the determination method of attribute weight,ranking method,two types of consensus reaching methods("individual-individual" and "individual-group")and the determination method of decision makers' weights,etc,which are used to solve the actual decision-making problem for the project evaluation of venture capital.The main research contents are shown as follows:(1)For the selection problem of the Multi-Attribute Decision Making under the environment of probabilistic linguistic term sets,a method based on the information entropy of the MAGDM with probabilistic linguistic is proposed.The measurement methods of probabilistic linguistic entropy and probabilistic linguistic cross entropy are respectively proposed.Then,by introducing the equilibrium factor,the measurement method of probabilistic linguistic combined entropy is proposed to solve the problem for the unknown attribute's weight of MAGDM.Furthermore,a new information aggregation method is proposed to obtain collective decision information,and according to the basic principle of VIKOR,a probabilistic linguistic group decision making method based on VIKOR is proposed to obtain the ranking result.The method not only considers the information and the relationship of attributes themselves,but also provides a new aggregation method for probabilistic linguistic term sets.In addition,the method also consider the psychological characteristics of decision makers' with different preferences when face with "maximum group benefit"and "minimum regret value',which is more satisfy to the actual decision-making problems.(2)For the consensus problem of "individual-individual" decision-making under the environment of probabilistic linguistic term sets an interactive consensus reaching method is proposed to improve the consensus degree.Firstly,a new distance measurement method of probabilistic linguistic term sets is proposed,which is used to measure the consensus degree in decision making.Then,the decision consensus degree threshold is extracted by several times of computer numerical calculations.Furthermore,for the "individual-individual" type of consensus problem in group decision making,an interactive algorithm based on consistent opinion is proposed to improve the consensus degree.This method can not only provide a new distance measurement method for probabilistic linguistic term sets,but also avoid subjectivity and empiricism in threshold determination.In terms of consensus problem of group decision making,consensus researching algorithm of group decision-making is constructed by incorporating consistency information,which can largely avoid the participation of inconsistent decision makers,so as to improve the reliability of adjustment opinions.(3)For the consensus problem of "individual-group" decision-making under the environment of probabilistic linguistic term sets,an adjustment method of group interaction consensus based on trust relationship is proposed.Firstly,in order to obtain the complete trust relationship,the probabilistic linguistic binary trust function and the trust transfer algorithm based on the strength of multi-path relationship are proposed.Next,a rule is proposed to adjust the trust relationship in order to gain the weight of decision makers'weights.Finally,for the "individual-group" type of consensus problem in group decision making,a consensus reaching algorithm of group interaction based on trust relationship is proposed.This method can integrate the social relations of decision makers into the consensus reaching algorithm for the group decision making,which can provide a more practical decision-making method.It not only considers the important social characteristics of decision-makers,but also more intuitively presents the dynamic process of group satisfaction consensus adjustment.(4)On the basis of the existing research of the project evaluation of venture capital,the project evaluation mechanism of venture capital under probability linguistic environment is analysized,and then the evaluation indices system and decision-making process for the project evaluation of venture capital are establisehed.Through the analysis of actual case,the application value of the methods proposed in thesis are illustrated by comprehensively.using the above methods of multi-attribute group decision making with probability linguistic term setsThe research results in this thesis can provide a new perspective for the method of MAGDM with probabilistic linguistic term sets,which can enrich and develop the theory and methods of MAGDM with linguistic,and provide new methods and tools for the project evaluation of venture capital project in practical.