Self-Organizing Maps for Large Databases: A Really Fast Approach

Carlos Bedregal (2), Ernesto Cuadros-Vargas (1,2)

e-mails: cbedregal@usp.edu.pe.com, ecuadros@spc.org.pe

(1) Sociedad Peruana de Computación
(2) Universidad Católica San Pablo

Abstract

Nowadays the need the process lots of complex multimedia databases is more frequent. Recent investigations such as MAM-SOMS* and SAM-SOMS* families propose the combination of Self-Organizing Maps (SOM) with Access Methods for a faster similarity information retrieval. In this paper we present the propierties and experimental result of the MAM-SOM* techniques that show and improvement of up to 90% using recent Access Methods such as Slim-Tree and Omni-Sequential in contrast with a traditional SOM network.

PDF de este artículo

PDF de JPC2006 (incluye todos los artículos)