Evolutionary Clustering as Technique of Economics Problems Solving

UDC 004.896 (045), DOI:10.18372/1990-5548.54.12333

Snytyuk V. Y., Suprun O. O. Evolutionary Clustering as Technique of Economics Problems Solving // Electronics and Control Systems, N 4(54) – Kyiv: NAU, 2017. – pp. 95–101.

The article presents the method, developed to use evolutionary technologies for clustering large amount of objects that are specified by their characteristics values. The need to analyze big data and to extract the necessary data from multidimensional databases makes the classic methods ineffective, or they require a lot of recourses or time to give an appropriate solution to the stated practical problem. Such problems very often appear in economical and financial spheres, where an expert has to make right decisions, based on various information from different sources, this information may have noise effects, or even be unreliable. Solving these problems requires gathering and formalization of available information that can take a lot of time. The presented method allows to use evolutionary technologies, such as genetic algorithms and evolution strategies elements to solve clustering problems with minimal constraints on the initial data – the situation that represents real practical problems. The experimental results of using the method are given, which proof the effectiveness of the proposed methods.

Index Terms—Clustering problem; complex objects; evolution technologies; genetic algorithm;
evolution strategies.

References: 14 name.