Anti-Entropy Resolving of Uncertainty of Estimations Within Scope of Intelligent DMSS

Abstract

Considered as conceptual, mathematical and algorithmic ways to resolve uncertainty that occurs sporadically when there is a finite set of estimations of a dynamic trajectory of a quantitative characteristic value of an arbitrary length when various DMSS types are functioning. The consideration was limited to a case of non-discrete characteristics, while assuming that the information about the values of the extent of indetermination of those estimations is a priori known. In this article is formulated and solved for a complex problem task for the resolving of a respective uncertainty, which is conceptually and instrumentally complicated by a rigorous requirement to increase the reliability of the resultant estimation. This article proves the conceptual universality of the proposed procedure for the resolving of uncertainties. A deductively generated hypothesis as to mechanical applicability of the method is used on a case of a countable set of estimations.

A hybrid MCDM model combining Fuzzy-Delphi, AEW, BWM, and MARCOS for digital economy development comprehensive evaluation of 31 provincial level regions in China

“…For AEW approach, assuming the amount of alternatives is m and the amount of sub-criteria is n , the value of sub-criteria i with regard to alternative j is x ij ( i = 1,2,…,n, j = 1,2,…,m), hence, the assessment matrix is described as X = [ x ij ] n×m . Thus, the anti-entropy value of every sub-criterion can be obtained via [ 33 ]: where r ij = x ij /∑ j x ij . Via standardizing the anti-entropy value, the objective weight w 1 i of every sub-criterion is calculated by: …”

“….,m), hence, the assessment matrix is described as X = [x ij ] n×m . Thus, the anti-entropy value of every sub-criterion can be obtained via [33]:…”

Development of System Typology and Choice of Preferred Simulation Modelling Methods for DSS-Toolkit

“…Some methods of mathematical modelling founded on reproduction of random events both processes and methods, referred to a category of simulation modelling (see fig. 1), now are known -see, for example, prototypes of description in Dmitriev, 2002a;Dmitriev, 2002b;Dmitriev, Dergunov, 2004;Dmitriev, 2011. The so-called methods of statistical tests take an intermediate position between analytical and simulative methods therefore they are detailed into the independent class.…”

“…Since any mathematical model of the managed object is used in the implementation of several, typical management functions, the problems of deviations according to the results of simulation experiments can increase cumulatively and, accordingly, synergistic effects of the appearance and exacerbation of errors are very likely to occur: see, for example, Dmitriev, 2002b. Therefore, the introduction of a system typology of the conceptual variety of simulation modelling methods and the ranking procedure for their preference seems to be a productive, creative measure, an exhaustive solution for which has not been identified in accessible sources.…”