Application of classification algorithms of Machine learning in cybersecurity
Gunay Abdiyeva-Aliyeva, Jeyhun Aliyev, Ulfat Sadigov · Procedia Computer Science · 2022
In the contemporary period, the cyber-crime activities have been one of the significant problems in the both private and public organizations. Building real-time checking systems to prevent digital crime activities is challenging but very useful in the detection of cyber-attacks both in time and effective. Machine Learning (ML) is a part of the data science where these mentioned problems could be solved with automated processes. Variety of algorithms in ML can be applied in cybersecurity to prevent attacks and minimize the security risks. Extreme Gradient Boosting (XGBoost) is one of those algorithms can be used effectively in the cybercrime detection activities. Due to the huge number of manual jobs in the investigation process and rule-based prevention methods, XGboost is thought to be automate those processes. In the paper, how ML algorithms can be applied in detection of cybercrime activities, the mathematical background in XGBoost and the main features selected to this problem are discussed.