Machine Learning in Cybersecurity: Systematic Literature Review

Eduardo Valdiviezo Salazar M.; Yeferson Huillca Rojas F.; Felipe Alarcón Vázquez S. · Proceedings of the LACCEI international Multi-conference for Engineering, Education and Technology (2024) · DOI: 10.18687/laccei2024.1.1.723

Resumen

Technological advancement has created an urgent demand to strengthen cybersecurity, facing incidents such as Denial of Service (DOS) and Distributed Denial of Service (DDOS) attacks. Thus, this systematic review aims to investigate current technologies in cybersecurity, to strengthen digital security and provide key information on the latest trends, contributing to the continuous adaptation of cybersecurity strategies. The methodology, based on the PICO strategy, structures the search in the Scopus and IEEE databases, selecting 21 publications out of a total of 308. The results highlight the complexity of cybersecurity and the variability in the effectiveness of machine learning algorithms, underscoring the importance of careful tool selection. In addition, it is observed that, on average, Decision Tree algorithms achieved 99.59% accuracy for DOS attacks, with a crucial role in defending against cyber threats. The conclusion highlights the critical need for adaptable strategies supported by efficiencies ranging from 41% to 99%, suggesting exploring hybrid approaches and emerging challenges to continuously strengthen cybersecurity. In addition, a 99.6% detection rate underscores the importance of choosing tools carefully, with 32% on false positives and 16% on metrics such as accuracy and recall, emphasizing the need for anticipation and flexibility for effective cybersecurity.

Instituciones: Universidad Tecnológica del Perú