2026-06-10T23:59:28.772Z http://ezaposleni.singidunum.ac.rs/rest/sciNaucniRezultati/oai

ezaposleni.singidunum.ac.rs/rest/sciNaucniRezultati/oai:3:9672 2023-11-18T16:14:03Z 3

Detection of BotNet Using Extreme Learning Machine Tuned by Enhanced Sine Cosine Algorithm, Chapter in LNEE Lecture Notes in Electrical Engineering: ICAAAIML 2022: Advances and Applications of Artificial Intelligence & Machine Learning, Springer, volume 1078 2023 http://ezaposleni.singidunum.ac.rs/rest/sciNaucniRezultati/oai/record/3/9672 https://link.springer.com/chapter/10.1007/978-981-99-5974-7_12 N. Bacanin M. Zivkovic Z. Hajdarevic A. Petrovic N. Budimirovic M. Antonijevic I. Strumberger The rise of the internet of things (IoT) popularity with smart home systems and improvements in the technologies of embedded devices raised interest in such solutions. However, with the increase in the utilization of these technologies, the security risks increased as well. Connecting a large number of devices to the same network is risky, because the security of any network is only as strong as its weakest part, and in the case of IoT the weakest link can be a small home appliance. This research aims to improve the means for detecting and preventing attacks on such networks via a hybrid method between machine learning and population-based metaheuristics. The proposed study introduces an enhanced version of the recently emerged sine cosine algorithm (SCA) that was used for tuning extreme learning machine (ELM). The hybrid method was validated against the UNSW-NB15 dataset that consists of features that describe regular (normal) and botnet traffic, where the latter refers to the distributed denial of service (DDoS) attack. The hybrid framework’s performance metrics were compared with several other classic machine learning models, as well as with other ELM solutions tuned by other population-based metaheuristics. Based on experimental data, the conclusion that the proposed method on average obtains the best performance can be derived. bookPart Springer, Singapore 125 137 10.1007/978-981-99-5974-7_12 LNEE Lecture Notes in Electrical Engineering: ICAAAIML 2022: Advances and Applications of Artificial Intelligence & Machine Learning, volume 1078