Alex L. Ramos, Cícero N. dos Santos.
Intrusion Detection is the process of monitoring events occurring in a network and analyzing them for signs of intrusion. The literature contains many papers that use ensemble of machine learning algorithms to solve detection problems. This paper proposes a model of detection in three levels. At each level, classification algorithms are applied and their results are combined in the later level. The last level consists of an ensemble of ensembles, which aims to improve the precision of intrusion detection. The results show that the use of three-level ensemble performs better than other systems proposed in the literature.
http://www.lbd.dcc.ufmg.br/colecoes/sbseg/2011/0016.pdfCaso o link acima esteja inválido, faça uma busca pelo texto completo na Web: Buscar na Web