Study On The Evaluation Method Of Preference Based On Fuzzy Clustering Iterative Model

The categories of data sample are evaluated efficiently, which is beneficial to improve the identifiability of data sample and help the data users make more reasonable and scientific decisions. But too subjective assessment methods may produce vastly different assessment results under different decision makers’ preference constraints. In addition, Different decision makers may carry out assessment work from the perspective of their own interests. Thus, the accuracy and reliability of the evaluation results are greatly reduced. And completely objective assessment results may be seriously deviated from the human cognition, Assessment results are difficult for the public to accept. Therefore, the decision maker preference and the sample data are effectively combined. It is very necessary to consider the evaluation method of the decision maker’s preference.This paper based on fuzzy clustering iteration model, from the point of view of decision maker’s preference, add the constraints of decision maker’s preference to the evaluation model. We put forward two kinds of comprehensive consideration of subjective and objective weight evaluation method. The assessment methods consider both the subjective preference of decision-makers and objective data distribution, which makes the evaluation results scientific and sufficient humanization. (1) This model introduces decision maker’s preferences into the differential evolution (DE) algorithm, to filter out those individuals which dissatisfy the preferences, after that, the assessment results that considered decision maker’s preferences could be obtained by optimizing the indexes’weight vector of the fuzzy clustering iterative model. (2) The subjective preference of decision makers add directly into the objective function of fuzzy clustering iteration model, the optimal solutions of objective function, namely, the assessment result of decision-makers.The preference of the decision maker response to the weight vector of attribute directly in the optimal solution, the optimal membership matrix is the result of the comprehensive consideration of the subjective and objective factors. This paper solves the comprehensive evaluation value of sample and the ranking of data sample, which is obtained by the combination of the optimal membership matrix and the weight vector of attribute. The case analysis shows that the sorting method of this paper is more reasonable than the type of the traditional sample sorting index category characteristic value index.At last, this paper bases on the evaluation methods of preference of the decision maker fuzzy clustering iterative model, which is applied to flood grade evaluation and assessment of an enterprise’s credit rating, and no preference assessment results are compared. This paper verify the evaluating method for the rationality, the validity and feasibility, which illustrates the significance of the decision maker’s preference in the evaluation work and the necessity of quotation the decision maker’s preference in the evaluation process.