Flavius L. Gorgônio,  José Alfredo F. Costa. 

Distributed databases have become more common and require decentralized data clustering algorithms. In many cases, traditional methods are not suitable. Additionally, security and privacy preserving questions frequently make impossible to integrate very large databases. In these cases, distributed alternative approaches are required. This work presents a strategy to perform clustering analysis in distributed databases using parallel self-organizing maps. Preliminary comparative tests shown that the distributed approach performed similar results to the centered one.

http://www.lbd.dcc.ufmg.br:8080/colecoes/waamd/2007/008.pdf

Processando consulta.