Clustering data using the growing neural gas method

Abstract

Clustering data using the growing neural gas method

Chernov A.V., Butakova M.A., Shevchuk P.S.

Incoming article date: 12.07.2020

The article discusses the problems that arise in pattern recognition related to clustering and data abstraction. Detailed typical data clustering options. The problem of data transformation by vector quantization with the least error is given. A competitive training system for an artificial neural network based on a growing neural gas is described. Using the method of growing neural gas, an improved algorithm of a self-learning artificial neural network of competitive training is proposed. The criteria for completing clustering using the adaptation criterion as a stop criterion are defined. Examples of data clustering by an artificial neural network using the method of growing neural gas are given.

Keywords: clustering, artificial neural network, computer modeling, pattern recognition, self-learning intelligent systems