Resumo

RIVERO PEREZ, Jorge Luis. Machine learning techniques for intrusion detection in computer networks. Rev cuba cienc informat [online]. 2014, vol.8, n.4, pp. 52-73. ISSN 2227-1899.

The development of network intrusion detection systems (NIDS) is a challenge for researchers, due to the growth of computer networks, constantly appear new content-based attacks. This article in addition to do a description of the approaches of intrusion signature-based and anomaly detection ones, also constitutes a review of the different machine learning techniques for the intrusion detection to be applied in data preprocessing and processing stages. NIDS taxonomy and an attributes classification scheme are described. In anomaly detection from machine learning techniques several data sets are employed, being KDD Cup 99 the most used. That data set is described and the results of some data preprocessing techniques applied on it such as selection and discretization are presented. Novel approaches that use search algorithms based on swarm intelligence with machine learning algorithms are exposed, which increase detection rates and improve the detection of content-based attacks. This review is of great relevance to researchers looking for areas within the intrusion detection in computer networks using machine learning techniques, in which make contributions.

Palavras-chave : intrusion detection; KDD Cup 99; machine learning; swarm intelligence.

        · resumo em Espanhol     · texto em Espanhol     · Espanhol ( pdf )