Mohammad, Abdul Rahman and Hasan, Mahbub and Reza, Jafrin and Bhuiyan, Mohd Salman Hossain (2024) AI-Driven Cybersecurity for Business: Predictive Detection of Malware and Network Intrusion. American Journal of Technology Advancement, 1 (1). ISSN 2997-9382
|
Text
AJTA_AI-Driven+Cybersecurity.pdf Download (595kB) |
Abstract
As businesses grow ever more reliant on digital technologies, interdependent networks and information systems, malware attacks, unauthorized access and network intrusions have grown more dangerous. The traditional security mechanisms might not be effective in dealing with the changes in the threats, especially if malicious activities are not following the patterns they are familiar with or if they are attempting to beat the security mechanisms with non-traditional approaches. Artificial intelligence (AI) specifically machine learning can be used to improve cyber security by detecting patterns in network traffic and system memory to determine if it is malicious or not. The application of AI driven techniques in business for predictive malware detection and network intrusion detection is discussed. The study is based on two publicly available cybersecurity datasets: CIC-MalMem-2022, which provides support for detecting malware using memory analysis approaches; and CICIoT2023, which includes both malicious and benign network traffic data for various intrusion classes. Random Forest and XGBoost are suggested as the baseline machine-learning algorithms to assess both detection tasks. The datasets will be analysed separately since they are used to solve different cybersecurity problems and are composed of different sets of features. The evaluation of the model performance will be done based on the accuracy, precision, recall, F1 score, false positive rate and the model efficiency in terms of computation, with special emphasis on class imbalance and detection reliability. The study will also examine the role of AI-based detection capabilities in a business' risk management, incident response, digital asset protection, and possible impact on operations of a cyber-attack. Furthermore, implementation issues such as the risk of false alarms, computational needs, model generalization and data quality will be analyzed. In this study, the researchers used a comprehensive research platform to examine malware and network intrusion detection, in order to offer evidence-based findings regarding the possibilities and challenges of AI in cybersecurity for companies. The results of the experiments will be used to make recommendations for embedding machine-learning detection models into an organization’s security plan and to highlight that simply being able to classify existing records doesn't mean they can be able to predict future attacks before malicious acts happen.
| Item Type: | Article |
|---|---|
| Uncontrolled Keywords: | Artificial Intelligence, Cybersecurity, Malware Detection, Network Intrusion Detection, Machine Learning and Business Risk Management |
| Subjects: | H Social Sciences |
| Depositing User: | admin eprints |
| Date Deposited: | 15 Nov 2026 08:27 |
| Last Modified: | 15 Nov 2026 08:27 |
| URI: | http://eprints.umsida.ac.id/id/eprint/17559 |
Actions (login required)
![]() |
View Item |
Dimensions
Dimensions