Software implementation of the main cluster analysis tools

• Silin, Andrey V. ^[2] ; Grinyuk, Olga N. ^[2] ; Lartseva, Tatyana A. ^[1] ; Aleksashina, Olga V. ^[1] ; Sukhova, Tatiana S. ^[3]
  1. [1] Moscow Polytechnic University

     Moscow Polytechnic University

     Rusia

     
  2. [2] Mendeleev University of Chemical Technology of Russia, Novomoskovsk, Russia.
  3. [3] Moscow Aviation Institute (National Research University), Moscow, Russia.

  Mostrar afiliaciones +
• Localización: Revista Amazonia Investiga, ISSN-e 2322-6307, Vol. 10, Nº. 47, 2021, págs. 81-92
• Idioma: inglés
• Enlaces
  • Texto completo (pdf)
• Resumen

  • This article discusses an approach to creating a complex of programs for the implementation of cluster analysis methods. A number of cluster analysis tools for processing the initial data set and their software implementation are analyzed, as well as the complexity of the application of cluster data analysis. An approach to data is generalized from the point of view of factual material that supplies information for the problem under study and is the basis for discussion, analysis and decision-making. Cluster analysis is a procedure that combines objects or variables into groups based on a given rule. The work provides a grouping of multivariate data using proximity measures such as sample correlation coefficient and its module, cosine of the angle between vectors and Euclidean distance. The authors proposed a method for grouping by centers, by the nearest neighbor and by selected standards.

    The results can be used by analysts in the process of creating a data analysis structure and will improve the efficiency of clustering algorithms. The practical significance of the results of the application of the developed algorithms is expressed in the software package created by means of the C ++ language in the VS environment.