Ανίχνευση και πρόβλεψη κυβερνοαπειλών : μελέτη περίπτωσης για κατανεμημένες επιθέσεις άρνησης υπηρεσίας
Cyber threat detection and prediction : a case study for distributed denial of service attacks
View/
Keywords
DDoS ; EWS ; Cyber threat prediction ; Cyber threat detectionAbstract
Detection and prediction of cyber threats are considered one of the most complicated challenges faced by cybersecurity teams in the era of rapid digital growth and instant online communication required by modern computing systems. In recent years, this field has engaged both incident-response and security-operation teams as well as the academic cybersecurity community. To this end, several efforts have been made to develop sophisticated systems for detecting and predicting cyberthreats to provide security analysts valuable time to address the potential incident. Moreover, the integration of data-science techniques has significantly improved the accuracy of these implementations by leveraging artificial intelligence and machine learning methodologies.
This thesis analyses several implementations for detecting and predicting cyberattacks and presents a comparative case study of tools focusing on predicting Distributed Denial of Service (DDoS) attacks. The purpose of this paper is to compare the different machine learning techniques used in the selected implementations in the aspects of model accuracy and time prediction.