Clustering as a tool for managing industrial enterprise

Abstract

CluSTerinG aS a Tool for ManaGinG inDuSTrial enTerPriSe Purpose. Substantiation of methodological foundations for forming a cluster of industrial enterprises and establishing a system of relationships between their cluster groups. Methodology. Specific analogue modelling techniques were used to identify the relations between manufacturing enterprises; economic and mathematical modelling was used to search for multilayered network communities. findings. A fundamentally new methodological basis has been proposed for identifying a cluster of industrial enterprises. This has been done through a comparison of the results of the three approaches to clustering. It has been revealed that hierarchical cluster analysis does not allow identifying similar groups of enterprises or relationships between them, since this approach lacks a single strict criterion for an optimal split of the dendrogram into clusters. The competitive approach of the geometric distance of neurons to objects, which is based on the technique of selflearning neural network and Kohonen's selforganizing maps, also identified a nonuniform cluster structure. The study proposes to form a cluster of industrial enterprises through the method of searching for communities in multilayered network graphs. This method was a breakthrough in building a cluster as a merger of extractive and processing industrial enterprises, together with academic and research institutions. originality. A new methodological approach to the formation of industrial enterprise cluster has been proposed, whose math ematical basis was developed by T. Kamada. This approach uses multiple object proximity matrices, which take into account sup plierconsumer relationships, geographical distances, and patterns of ownership. It has been proved that this method for clustering is more advantageous, since it allows identifying the communities of enterprises, which are network analogues to a cluster; it also takes into account the relationships of the analysed metallurgical enterprises of the mining and processing industry with educa tional and research institutions of the enterprises. The development of these relationships creates the basis for the productive de velopment, efficient operation and additional competitive advantages for industrial enterprises. Practical value. Under conditions of a crisis in the metallurgical industry, it is recommended to create a cluster, which will significantly increase the competitiveness of each enterprise included in the cluster and fully use the potential of the metallurgical complex.

Economic security of regions

“…The scientists have proven, and we agree, that economic-mathematical modeling is a productive methodological approach to assessing economic security (Cable 1995, Granberg et al 2008, Kharazishvili 2019, Iefimova et al 2020, Chichkanov et al 2020, Onyshchenko, Bondarevska 2018. Thus, significant advances in assessing the level of economic security of regions were achieved by using the clustering method (Ivanova et al 2020, 2021, Ovcharenko et al 2021) and the nonlinear approach and the method of fuzzy logic, whose current methodological foundations were substantiated by Piegat (2001), Zelentsov, Korotka (2018). In view of this, the authors of this article in their previous research used the method of clustering to assess the economic security of regions.…”

Managing Economic Security of Metallurgical Enterprises under the Conditions of Transformation Changes in the Economy

“…Today, the strategic goal of transformational changes in the economy is the formation of competitive industrial enterprises as equal partners in the world economic community, or clustered as a separate form of operation according to M. Ivanova [3]. The implementation of these requirements, as evidenced by world experience, is possible on the basis of consistent application of the tools of economic security management in enterprises, which should be based on an assessment of their capability to adapt in the changing space of functioning according to R. Miśkiewicz and R. Wolniak [11].…”