MACHINE LEARNING-BASED MITIGATING CYBER THREAT PREDICTION FOR STRENGTHENING SECURITY ACROSS CYBER SUPPLY CHAINS

Abstract

Cyber supply chains are complex systems that can be jammed by cyber threats affecting
the business. They need CTI to keep threat actor behavior, Tactics, Techniques, and
Procedures , and Indicators of Compromise under study and thus derive possible threats
to minimize them. This paper incorporates CTI characteristics into a ML-based Cyber
Threat Prediction model. Classification algorithms are used to predict cyber threats on
the Microsoft Malware Prediction dataset. The research focuses on creating models for
predicting advanced persistent threats, command-and-control exploit attacks, and
industrial espionage. Experiment results show that LG, SVM does well at an accuracy
level of 85%. The predictive analytics validated herein using ML will be used for
revealing vulnerabilities that enhance the security of cyber supply chains. It recommends
countermeasures for new threats like ransomware and spear-phishing in enhancing
proactive defense mechanisms.