Department of Mathematics, Faculty of Science, Khon Kaen University, Khon Kaen, Thailand, 40002.

Abstract

Due to the rapid increase of size and resolution of industrial and scientific data sets, artificial neural networkshave become essential tools for identifying the important aspects of the clustered data structure. In this article, wereviewed unsupervised neural network methods which can be applied to the task of extracting hidden structuresas useful features for subsequent processing. The unsupervised learning algorithms reviewed here are groupedinto two sections: data clustering methods and dimensionality reduction methods. Each of these major sectionsconcludes with a discussion of successful applications of the methods to image segmentation and data visualization.

Keywords : Unsupervised neural networks, Data clustering, Dimensionality reduction, Data visualization