Evaluating Features Selection on NSL-KDD Data-Set to Train a Support Vector Machine-Based Intrusion Detection System

The integrity of information and services is one of the more evident concerns in the world of global information security, due to the fact that it has economic repercussions on the digital industry. For this reason, big companies spend a lot of money on systems that protect them against cyber-attacks like Denial of Service attacks. In this article, we will use all the attributes of the data-set NSL-KDD to train and test a Support Vector Machine model. This model will then be applied to a method of feature selection to obtain the most relevant attributes within the aforementioned data-set and train the model again. The main goal is comparing the results obtained in both instances of training and validate which was more efficient. © 2019 IEEE.

• Fecha de publicación: 2019
• Institución: Universidad Tecnológica de Bolívar
• Repositorio: Repositorio Institucional UTB
• Idioma: eng
• OAI Identifier: oai:repositorio.utb.edu.co:20.500.12585/9137
• Acceso en línea: https://hdl.handle.net/20.500.12585/9137
• Palabra clave: Classification model; Data set; Dos Attacks; Feature selection; Machine learning; Support vector machine; Artificial intelligence; Classification (of information); Denial-of-service attack; Intrusion detection; Learning systems; Network security; Statistical tests; Support vector machines; Classification models; Cyber-attacks; Data set; Features selection; Intrusion Detection Systems; Support vector machine models; Feature extraction
• Rights: restrictedAccess
• License: http://creativecommons.org/licenses/by-nc-nd/4.0/