Scientific journal

ISSN 1814-2400

INFORMATION SCIENCE AND CONTROL SYSTEMS

Kazakovtsev L. A., Rozhnov I. P.

AN APPROACH TO THE DEVELOPMENT OF CLUSTERING ALGORITHMS BASED ON PARAMETRIC OPTIMIZATION MODELS

A new approach to the development of automatic grouping algorithms (clustering) based on parametric optimization models with the combined use of search algorithms with variable randomized neighborhoods and greedy agglomerative heuristic procedures is proposed. The experiments show that the new algorithms, developed using the new approach, make it possible to obtain more accurate and stable results (according to the achieved value of the objective function) in comparison with the known clustering algorithms for a fixed runtime, which allows using these algorithms in the interactive decision-making mode to solve practical problems.

Keywords: clustering, automatic grouping, k-means, k-medoid, CEM-algorithm