Fuzzy Clusterization of Ecological Risks in Geo-Information Systems

Abstract

An algorithm for solving the problem of fuzzy classification of environmental risks for practical use as part of the mathematical and algorithmic support of a geo-information automated system for the environmental risks' analysis and monitoring in the context of initial data fuzziness is considered. The analysis and recommendations on the choice of an adequate method for the fuzzy problem solution of environmental risks' automatic classification are made. It is proposed to use the Fuzzy C-Means approach to solve the represented problem of environmental risks' fuzzy classification. A model for the process of classifying ecological risks by the method of fuzzy clusterization is proposed and the problems that the experimental program module should solve are identified. An algorithm for solving the problem of fuzzy clusterization of environmental risks is proposed. A software module was implemented in the form of a special automated procedure for fuzzy clusterization used in the model GIS to determine the rational composition of environmental risks at the risk monitoring stage. The perspective directions of further research were evaluated with the aim of solving a wide range of optimization tasks for fuzzy identification, fuzzy analysis, fuzzy monitoring and fuzzy classification of the composition of controlled environmental risks.