Сentroid detection algorithm for data arrays in IoT paradigm

DOI №______

Abstract

The main algorithms of clustering and approaches to solving cluster analysis problems are considered. The analysis of actual problems
of cluster analysis is carried out. Disassembled the popular algorithm of K-means and its main advantages and disadvantages. The use
of the improved K-means algorithm is proposed and the effectiveness of this method is substantiated.

Keywords: algorithm; analysis; processing; identification; clustering.

References
1. Gowda, K. C. Agglomerative clustering using the concept of mutual nearest neighborhood / K. C. Gowda, G. Krishna // Pattern Recognition.— 1977. — V. 10.— P. 105–112.
2. Jain, A. K. Artificial neural networks: A tutorial / A. K. Jain // IEEE Computer.— 1996.— V. 29.— P. 31–44.
3. Zahn, C. T. Graph-theoretical methods for detecting and describing gestalt clusters / C. T. Zahn // IEEE Trans. Comput.— 1971.— C-20.— P. 68–86.
4. Загоруйко, Н. Г. Прикладные методы анализа данных и знаний / Н. Г. Загоруйко.— М., 1999.
5. Котов, А. Кластеризація даних / А. Котов, Н. Красильников.— К., 2006.