Research On Multi-criteria Sorting Group Decision Making Based On Case-based Reasoning

With the increasing complexity of decision-making tasks and more convenient of people’sinformation exchange, Multi-criteria Sorting Group Decision Making is becoming the focus of thedecision problem. Multi-criteria Sorting Group Decision Making is an important aspect of theMulti-criteria Group Decision Making theory, and it’s also the focus of current research. Due to thecomplexity of the issue of the actual decision-making and policy makers cognitive limitations,compared with specific information which is directly provided classification parameters, it makes thedecision-makers classified some programs or historical data easier to implement. Therefore Based onCase-Based Reasoning Multi-criteria Sorting Group Decision Making is more practical. However,how to extracted the information of the classification parameters from the Existing cases classifiedinformation is the key to such a classification problems. This thesis applies the evidence theory andthe gray clustering assessment to research the Multi-criteria Sorting Group Decision Making problembased on Case-Based Reasoning.Firstly, the thesis research the build-up of the Multi-criteria evaluation information under theGroup Decision Making, use the probability assignment function in the Evidence Theory to expressthe evaluation information, and use the rules of evidence synthesis to build them up. Then, we can geta sorting information about the build-up existing cases; Secondly, the thesis build aMulti-criteria evaluation information weight model which is based on gray entropy, and verify theclassification parameters, then use them to solve the Multi-criteria Sorting Group Decision Makingthrough the Gray-clustering evaluation Method. Main contents are as follows:(1)Apply the probability assignment function in the evidence theory to express uncertaintyevaluation information by a number of experts, analyze characteristics and shortcomings of the rulesof evidence synthesis, aimed at questions about evidence in the process of Group Decision-makinginformation assembled, we put forward a method that use grey correlation degree to estimate evidencesimilarity, then, amend the basic probability assignment of the relevant evidence. To concentrate theexperts’ evaluation information based on the average fusion rule.(2)About the Sorting Decision Making problem which is unknown weight and based on cases,first of all, analyze potential cases, expand the range of indicators, according to the characteristics ofindicators to determine suitable form of whitening weight function，then, build a certain weight modelbased on minimum gray entropy, according to the classification information and the relationship between the known index weights, extract weight parameter information in the classification process,finally, use Grey clustering to evaluate the objects with evaluation.(3)In order to verify the practicality of these build-up models, I did some research about theMulti-criteria evaluation for Government low-carbon policy performance. Construct the evaluationindex system on the basis of analyze the background and necessity of Government low-carbon policyperformance. Select some representative provinces and cities, collect the data of each index, Let theexperts evaluate these provinces and cities, and then build up the information by synthesis of thetwo-stage. Then, extract weight information from a certain weight model based on minimum grayentropy, use the information in the performance evaluation for other provinces and cities.