The algorithm for constructing membership functions by statistical processing of expert estimates

An algorithm for constructing membership functions by statistical processing of expert estimations of the distribution of ele-ments on the fuzzy set of fuzzy subsets of the set level approach. A characteristic feature of the developed algorithm is to re-place the expert facing challenging enough elements of finite groups by a clear plurality intersecting subsets level easier task distribution on interval subsets. The efficiency and effectiveness of the proposed algorithm is confirmed by comparative modeling. As the object of comparison chosen statistical method of estimation of fuzzy subsets of a finite set, based on subsets of level as a comparable parameter - deviation of the estimates obtained from the reference value . Comparing estimates obtained on the basis of the developed algorithm and the object of comparison showed that the estimate obtained with the help of the developed algorithm was closer to than the estimate obtained for the object of comparison by 16%. The scope of application of the developed algorithm is a system of decision-making on the basis of obeying Possibilistic extent to which the requirement of reliability of classification dominates the requirement to build-time membership function.