Open AccessKLASIFIKASI TINGKAT KEPARAHAN SERANGAN JARINGAN KOMPUTER DENGAN METODE MACHINE LEARNING
Author(s) -
Okki Setyawan,
Angge Firizkiansah,
Ahmad Nuryanto
Publication year - 2021
Publication title -
jisicom (journal of information system, infomatics and computing)
Language(s) - English
Resource type - Journals
eISSN - 2597-3673
pISSN - 2579-5201
DOI - 10.52362/jisicom.v5i1.443
Subject(s) - firewall (physics) , computer science , the internet , seriousness , computer security , artificial intelligence , machine learning , operating system , political science , law , charged black hole , physics , schwarzschild radius , classical mechanics , gravitation
Computer networks are currently developing very rapidly, so that many electronic devices are connected to the internet, but the security system adopted by these devices must be qualified so they are not vulnerable to threats and dangers. Researchers want to find out how severe the threat of an attack is detected by a firewall using data records from a company, using machine learning, namely K-Nearest Neighbors, Decission Tree. Classification of the severity of a computer network security system is usually called the severity level. In this study, the limitation of the seriousness level of the attack was divided into 3 parts from the highest level, namely critical, high and medium. The processed dataset is logging into the firewall as many as 5999 with 23 columns or features. The best of the three methods are K-Nearest Neighbors getting 100% accuracy and Decission Tree getting 100% accuracy . With the results of this data processing, the machine learning method is very suitable to be used to classify the severity of computer network attacks