ISSN 2079-3537      

 
 
 
                                                                                                                                                                                                                                                                                                                                                                                                                                                                             
Scientific Visualization
Issue Year: 2015
Quarter: 4
Volume: 7
Number: 5
Pages: 87 - 101
Article Name: METHODS DESIGN FOR VISUAL ANALYSIS OF CLUSTERS IN MULTIDIMENSIONAL DATA VOLUMES
Authors: A.E. Bondarev (Russian Federation), V.A. Galaktionov (Russian Federation)
Address: A.E. Bondarev
Keldysh Institute of Applied Mathematics RAS, Moscow, Russian Federation

V.A. Galaktionov
Keldysh Institute of Applied Mathematics RAS, Moscow, Russian Federation
Abstract: The paper considers design of algorithms intended for visual analysis of cluster structures in multidimensional data volumes. The paper is aimed to design of a set of visualization and visual analytics methods for cluster structure studies without applying of clusterization methods influencing at original data. To analyze clusters in original data volume we propose to use the methods of original data points mapping to enclosed manifolds having less dimensionality. The proposed approach is based on self-organized maps (SOM) design, principal components analysis (PCA) and application of elastic maps with further varying of elasticity parameters for the last ones. To provide complete processing of original data volume all mentioned above methods and approaches should be organized in a form of pipeline. The applying of such pipeline allows one to get insight of cluster structures at the different levels of details for multidimensional data volume in question.
Language: Russian


Open Article   Download ZIP archive