Model and Method of Processing Partial Estimates During Intelligent Data Processing Based on Fuzzy Measure

摘要

To substantiate the choice of the convolution algorithm for partial expert estimates, models of expert evaluation and the estimates convolution process to obtain an integrated index for each of the alternatives have been developed. These models, as well as the classification of convolution algorithms and the measures used, are developed on the basis of an ontological approach to the analysis of the subject area. The analysis results are verified using the mathematical apparatus of category theory. The construction of categorical-ontological models allowed to justify the use of a fuzzy measure of expert preferences to take into account the uncertainty manifested in the evaluation of alternatives. Improvements in the method of convolving partial expert estimates using a fuzzy integral have made it possible to develop and implement a software component for intelligent data processing. Its operation in production has increased the accuracy of the alternatives evaluation by expert groups and the productivity of the examination.