The Method of Clustering Geoinformation Data for Stationary Sectoral Geoinformation Systems Using Swarm Intelligence Methods

Abstract

It has been noted in the research that classical methods are not suf ficiently effective for finding optimal solutions in industry-specific geographic information systems (IGIS) because they do not consider logical parameters. It has been proposed to use clustering methods with the use of a conceptual model for selecting an optimization method, based on prior research of the objective function for a specific task. Clustering of spatial data is often required in IGIS. However, the objective function in such clustering typically exhibits multimodality, high dimensionality, and complex topology of the feasible value domain. Therefore, research has been conducted to select a clustering method in IGIS. The proposed method for selecting a swarm algorithm and parameters for clustering is based on the proposed quality and efficiency indicators of swarm algorithms. A method and basic information technology have been developed for clustering geoinformation data, where data are grouped into clusters based on compactness criteria using swarm methods, utilizing both quantitative and logical features, thus improving the clustering quality. A clustering method has been proposed that contributes to enhancing the quality of clustering, and the choice of swarm optimization method was made according to the developed methodology. Research on the efficiency of clustering using swarm intelligence methods has been conducted.