Hoque, Mohd Jahidul and Mohammad, Abdul Rahman and Hasan, Mahbub and Nafees, Muhammad Farhan (2023) Predictive Analytics and Machine Learning for Cyber Risk Assessment in Critical Information Systems: A Framework for Proactive Threat Identification and Risk Mitigation. American Journal of Economics and Business Management, 6 (3). ISSN 2576-5973

Text AJEBM_Predictive+Analytics+and+Machine+Learning.pdf Download (645kB)

Abstract

Based on historical data of cyber-attacks on critical information systems, this research examines the possibility of using predictive analytics and machine learning as proactive cyber risk assessment tools. The research uses exploratory data analysis to explore the extent of key types of cyberattacks, patterns of cyber escalation over time, regional differences in cyber risk distribution and the relationship between mitigation technologies and new threats. Results show that a limited number of attack types, including Malware, DDoS and Phishing attacks, account for most of the attacks worldwide, indicating the need to prioritize defensive resources. The temporal analysis also shows that the attacks have consistently increased over the years, culminating in a dramatic increase over the last few years, highlighting the predictive power of historical data in forecasting future attacks. The findings of geospatial analysis revealed inequitable distribution of cyber-risk, with technologically advanced countries having more cyber-attacks that are more frequent, and the emphasis on mitigation technologies also focused on vulnerability management and new solutions like blockchain and encryption. Predictive analytics and machine learning offer a methodological approach to pattern recognition, risk factor identification, and the ability to proactively identify threats. This framework is an improvement on cybersecurity planning because it integrates predictive threat detection and proactive mitigation measures, providing a structured way for organizations to plan for threats and allocate resources appropriately. The results will advance academic research and practical cybersecurity solutions by highlighting the potential of leveraging historical cyberattack data to translate it into actionable intelligence for proactive risk management in critical information systems.

Item Type:	Article
Uncontrolled Keywords:	Cyber-attacks, Information, Predictive Analytics, Machine Learning, Cyber Risk, Data
Subjects:	H Social Sciences
Depositing User:	admin eprints
Date Deposited:	31 Oct 2026 11:31
Last Modified:	31 Oct 2026 11:31
URI:	http://eprints.umsida.ac.id/id/eprint/17369

Actions (login required)

View Item