Abdullah, Chowdhury Amin and Ridoy, Md Jahidul Islam (2023) Artificial Intelligence-Driven Cyber Threat Detection for Protecting Critical Digital Infrastructure. Innovative: International Multidisciplinary Journal of Applied Technology (2995-486X), 1 (2). ISSN 2995-486X

Text Innovative_3931_Chowdhury+Amin+Abdullah_Artificial+Intelligence-Driven.pdf Download (501kB)

Abstract

Background: Energy grids and healthcare systems and financial networks and transportation systems more vulnerable to sophisticated cyber threats. The signature-based detection systems which rule-based cybersecurity methods use no longer provide sufficient protection against advanced persistent threats and zero-day exploits and AI-driven cyberattacks. Methods: We conducted their investigation through a data-based system which used simulated data that demonstrated actual conditions of essential infrastructure systems which included energy and healthcare and financial institutions. The dataset includes 500 entries which contain network traffic information and system operation patterns and transaction data elements that show both regular and harmful system activities. Random Forest and Gradient Boosting and Long Short-Term Memory (LSTM) and Auto encoder models through their performance on accuracy and precision and recall and F1-score and detection latency and error rates. Result: This finding demonstrate that AI-powered systems deliver better results than conventional methods because they produce accuracy percentages which range from 92% to 97%. Auto encoder model which uses deep learning techniques achieved the best results with 96.7% accuracy and maintained a false positive rate of 3.5%. The system achieved threat detection times which dropped by more than 90% to provide organizations with the ability to detect threats in real time. Conclusion: The models demonstrated outstanding performance when they needed to identify complex cyber threats included advanced persistent threats and zero-day attacks.

Item Type:	Article
Uncontrolled Keywords:	Cyber Threat, Artificial Intelligence, Critical Infrastructure, Threat Detection, Cyberattacks
Subjects:	H Social Sciences
Depositing User:	admin eprints
Date Deposited:	02 Jul 2026 02:48
Last Modified:	02 Jul 2026 02:48
URI:	http://eprints.umsida.ac.id/id/eprint/16721

Actions (login required)

View Item