DefinePK

Abstract

Many real-world problems compact with clustering of high-dimensional data, such as images, videos, text and web documents, DNA microarray data, and etc., frequently, such outside elements are clustered in the high dimensional data not addressed. In this paper, we propose an algorithm, called Local Adaptive Receptive Field Dimension Selective Self-Organizing Map of outsider in high dimensional data (LARFDSSOMOH), to cluster the data points that are in a high-dimensional subspaces and also cluster the outsiders (outside elements) that are available in the high dimensional data. The proposed mapping scheme enhances the system efficiency, by providing better quality of clustering when compared to its conventional counterpart. Finally, we explain the capability of the proposed algorithm through experiments on unnatural data as well as the natural data.