Research On Fuzzy Multi-attribute Decision Making Method Based On Association Rule Mining Algorithm

For the fuzzy multi-attribute decision making problem with historical decision data,the existing decision methods only utilize the information in the current decision data when making decisions,ignoring the useful information that may be hidden in the historical decision data,resulting in low credibility in the final decision results.To solve this problem,this paper uses the association rule mining algorithm to dig out the hidden useful information from the historical decision data,and applies the information in the process of fuzzy decision making to improve the credibility of the decision results.The main research contents of this paper are as follows:1.Aiming at the problem that the existing weighting methods do not utilize the historical decision data and the weight obtained is not reliable,a triangular fuzzy number weighting method based on association rules(WBARM)is proposed.This method finds out the attribute influence factors from historical decision data by association rule mining algorithm,and then integrates the attribute influence factors and the similarity difference index in the current decision data to determine the attribute weights,and the final obtained attribute weights are more reasonable.Aiming at the fuzzy multi-attribute decision making problem where the attribute weight is unknown and the attribute value is given in the form of triangular fuzzy number,a triangular fuzzy number multi-attribute decision making method based on WBARM is proposed.In this method weight of each attribute is assigned through WBARM,and the concept of possibility is introduced,then constructs the possibility comparison matrix to rank the alternatives.This method effectively uses the current decision data and the historical decision data in the decision-making process,and the decision-making results are more reliable.2.A fuzzy multi-attribute decision making method based on subjective and objective preferences is proposed for the fuzzy multi-attribute decision making problem where the attribute weights are not completely determined and the decision makers have preferences for alternatives.Firstly,aiming at the problem that the determination of subjective preference relies too much on the personal experience and knowledge of decision-makers,which is easy to cause the decision result is not consistent with the reality,a method of determining objective preference based on association rule mining is proposed.Then,the subjective and objective preferences are considered comprehensively,and the concept of similarity and possibility ranking method are introduced to construct the decision model,and then the model is used to prioritize each scheme.This method not only takes into account the subjective preferences of decision makers,but also makes effective use of objective decision data to ensure the reliability of decision results.3.A reference point determination method based on association rule mining(RPBARM)is proposed for the problem that the existing reference points in the cumulative prospect theory are not suitable as the gain or loss criteria of the scheme.This method uses association rule mining algorithm to find the reference points of the scheme from the historical decision data and the selected reference points can better reflect the actual gain or loss of the alternative scheme.Aiming at the problem of fuzzy multi-attribute decision making with unknown attribute weight and the decision maker has certain expectations for the attribute values,a cumulative prospect theory fuzzy multi-attribute decision making method based on RPBARM is proposed.Firstly,RPBARM is used to determine the scheme reference points.Then,the relative closeness of each alternative scheme is calculated by constructing the prospect gain value matrix and prospect loss value matrix,and then the order of each alternative scheme is determined according to the relative closeness.Finally,the proposed decision method is applied to different application scenarios of automobile purchase choice problem,and is compared and analyzed with existing fuzzy multi-attribute decision making methods of the same type to verify the feasibility and effectiveness of the proposed method.