Study on Machine Learning Techniques for Botnet Detection

• Autor(a) principal: Silva, L. [UNESP]
• Data de Publicação: 2020
• Outros Autores: Utimura, L. [UNESP], Costa, K. [UNESP], Silva, M. [UNESP], Prado, S. [UNESP]
• Tipo de documento: Artigo
• Idioma: eng
• Título da fonte: Repositório Institucional da UNESP
• Texto Completo: http://dx.doi.org/10.1109/TLA.2020.9082916; http://hdl.handle.net/11449/195369
• topic: Botnet; Machine Learning; Recursive Feature Elimination

This paper presents a study on the application of machine learning techniques for botnet detection, compromised computer networks controlled by an attacker in order to perform malicious activities, such as distributed denial-of-service attacks (DDoS), data theft and others. The study aims to evaluate the efficiency of commonly used classifiers in the literature for botnet traffic classification and, to this end, we compare the results obtained from each classifier using two different approaches for feature selection, the first one taking into account the most frequently used features in problems of this nature, based on previous works, and the second one taking into account features selected by the Recursive Feature Elimination algorithm, a relatively unexplored feature selection method in the botnet detection area.