РЕЗУЛЬТАТИ КЛАСТЕРИЗАЦІЇ ДІЛЯНОК РІЗНОЇ ПОТУЖНОСТІ ВУГІЛЬНОГО ПЛАСТА С10В ШАХТИ «ДНІПРОВСЬКА» ЗА ВМІСТОМ ГЕРМАНІЮ

Abstract

Purpose. The purpose of the work is to establish the most effective method of creating an objective typification of sections of coal seam с10в of the “Dniprovska” mine of different thickness according to germanium concentrations, based on the analysis of the clustering results. Methodology. Typification procedure is the systematization of objects according to a priori given features. Cluster analysis, taxonomy, pattern recognition, and factor analysis are usually used for this purpose. To achieve the goal set in the work, in the process of research, clustering was carried out using various methods, which are implemented in the most popular professional statistical software platforms “STATISTICA” and “SPSS”; their analysis was performed and the choice of the most optimal of them was substantiated. Excel 2016, STATISTICA 13.3 and IBM SPSS Statistics 22 versions were used in the work. Results. The main results of the research consist in the selection of the optimal method of clustering of areas of different thickness of the coal seam. The analysis of the dendrogram of the results of clustering by the weighted centroid median method of the с10в seam sections by germanium content, unlike others, allows not only to achieve the most stable division of the entire set of sections under consideration, but also to maximize the visualization of their breakdown by classes at different scale levels in the absence of a priori hypotheses regarding the number clusters and their forms. At the same time, the structure of clusters is clearly distinguished, regardless of the scale level of their formation, and the sequence of combining individual deposits and their groups into the resulting cluster is clearly traced. These advantages make it possible to make maximum use of already existing information for the development of natural typifications of areas of the с10в coal seam by germanium content and to interpret the obtained results in geological terms. Scientific novelty. The scientific novelty of the results of the conducted research consists in the establishment of the weighted centroid median method of cluster analysis, which is implemented in the professional statistical software platforms “STATISTICA” and “SPSS” as the most optimal for the subjectivity-free researcher of the division of sections of the coal seam с10в of the “Dniprovska” mine by content germanium into taxa. Practical significance. The practical significance of the work results is that the constructed dendrograms of the clustering of deposits by germanium content can be used as a basis for the development of a natural typification of the coal seams of the Dniprovska mine for their subsequent geological and economic assessment. This, in turn, will make it possible to make maximum use of already available information and interpret the obtained results in geological and genetic concepts, which will provide the opportunity to use it for the comprehensive use of mineral raw materials and to solve strategic issues of sustainable development of Ukraine. Key words: germanium, cluster analysis, coal seam, taxa, clusters, weighted centroid method, dendrogram of clustering results.