ISSN 1991-2927
 

ACP № 4 (58) 2019

Methodology and the Algorithm for Clustering Economic Analytics Object

Pavel Vladimirovich Dudarin, Ulyanovsk State Technical University, Postgraduate Student at the Department of Information Systems of Ulyanovsk State Technical University (UlSTU); graduated from the Ulyanovsk State Technical University; an author of papers in the field of text clustering. [e-mail: This email address is being protected from spambots. You need JavaScript enabled to view it. ]P. Dudarin,

Aleksandr Petrovich Pinkov, Ulyanovsk State Technical University, Candidate of Economics, graduated from Ulyanovsk branch of Kuibyshev Planning Institute; Acting Rector of Ulyanovsk State Technical University, an author of more than 50 papers, a monograph, and articles in the field of economics, planning, marketing, production engineering, higher education organization, and corporate training. [e-mail: This email address is being protected from spambots. You need JavaScript enabled to view it. ]A. Pinkov,

Nadezhda Glebovna Yarushkina, Ulyanovsk State Technical University, Doctor of Engineering, Professor; First Vice-Rector - Vice-Rector for Scientific Affairs of Ulyanovsk State Technical University; Head of the Department of Information Systems of UlSTU; graduated from Ulyanovsk Polytechnic Institute; an author of more than 300 papers in the field of soft computing, fuzzy logic, and hybrid systems. [e-mail: This email address is being protected from spambots. You need JavaScript enabled to view it. ]N. Yarushkina

Methodology and the Algorithm for Clustering Economic Analytics Object 000_12.pdf

The purpose of the study, the results of which are described in the article, consist in developing new and modified methods and algorithms for solving the problem of clustering objects of economic analytics. The use of known algorithms for clustering formulations of economic indicators in order to determine the similarity of objects is complicated by the fact that the formulation of indicators are very short and the traditional indicators of terms occurrence (frequency) are inadequate. In addition, widespread occurrence of interviews and various forms of questionnaires in economic analysis implies the use of linguistic estimates. For example, "customer satisfaction level" indicator is difficult to quantify, so, instead of conventional points, fuzzy values such as “high”, “medium”, “low” are often used. As a result, it becomes feasible to use a fuzzy variant of the k-means method - the method of fuzzy k-means. Typically, the number of indicators in economic analysis is quite big, which makes it advisable to modify the algorithm on the basis of parallel execution. The study addresses the following issues: the k-means method is modified, it is adapted to the characteristics of the economic analytics objects; the methodology of data preprocessing for clustering is developed; new versions of clustering objects of economic analytics are developed, and experimental research of the effectiveness of the developed methods for large volumes of data is carried out.

Fcm-алгоритм, clustering, method of k-means, economic analysis, big data, fcm-algorithm, parallelization.

© FRPC JSC 'RPA 'Mars', 2009-2018 The web-site runs on Joomla!