Machine Learning Enhanced Intrusion Detection and Anomaly Classification for MIS-Driven Enterprise Network Security
International Journal of Scientific Interdisciplinary Research · 01.09.2024 · 14 (09) · 01-41
DOI: 10.63125/xk6kpc72 ↗Kurz gefasst
Fachartikel in der Zeitschrift International Journal of Scientific Interdisciplinary Research, online erschienen am 01.09.2024. Verfasst von Kynatun Mahin und 1 weiteren Autor:innen. Die Arbeit wurde bisher 5-mal zitiert (Crossref).
Abstract (Originalsprache)
This study addresses the persistent challenge of protecting MIS-driven enterprise networks against increasingly complex, previously unseen, and adaptive cyber threats when conventional signature-based controls, fragmented security data, and static monitoring practices provide limited visibility and response capability. The purpose was to examine how Machine Learning-Based Intrusion Detection Capability (MLIDC), Intelligent Anomaly Classification Capability (IACC), MIS-Security Data Integration (MSDI), and Adaptive Threat Monitoring and Response (ATMR) influence Enterprise Network Security Performance (ENSP) across enterprise environments containing interconnected digital, network, and cloud-enabled resources. A quantitative, cross-sectional, explanatory, case-based survey design was adopted using purposive sampling of cybersecurity analysts, network administrators, IT/MIS’S managers, security engineers, system administrators, and IT risk or compliance professionals. Of 320 questionnaires distributed, 298 were returned and 286 valid responses were retained, yielding an 89.4% valid response rate. Data were analyzed using descriptive statistics, Cronbach’s alpha, KMO and Bartlett tests, Pearson correlation, multiple regression, ANOVA, multicollinearity diagnostics, and residual testing. ENSP recorded the highest mean (M = 4.21, SD = 0.51), followed by IACC (M = 4.18, SD = 0.54), ATMR (M = 4.16, SD = 0.56), MLIDC (M = 4.13, SD = 0.58), and MSDI (M = 4.09, SD = 0.60). All predictors were positively associated with ENSP: IACC (r = .78), ATMR (r = .75), MSDI (r = .71), and MLIDC (r = .69), all p < .001. The regression model explained 71.4% of ENSP variance (R² = .714, adjusted R² = .710, F (4,281) = 175.40, p < .001), with IACC the strongest predictor (β = .34), followed by ATMR (β = .28), MSDI (β = .23), and MLIDC (β = .20). The findings imply that enterprise cybersecurity improves when intelligent detection is integrated with accurate anomaly classification, coordinated security information, and adaptive threat response, supporting security operations, MIS visibility, faster incident handling, and resilient enterprise network protection.
Zitieren
APA
Mahin, K., & Ashfaq, S. (2024). Machine Learning Enhanced Intrusion Detection and Anomaly Classification for MIS-Driven Enterprise Network Security. International Journal of Scientific Interdisciplinary Research, 14(09), 01-41. https://doi.org/10.63125/xk6kpc72
Vancouver
Mahin K, Ashfaq S. Machine Learning Enhanced Intrusion Detection and Anomaly Classification for MIS-Driven Enterprise Network Security. International Journal of Scientific Interdisciplinary Research. 2024;14(09):01-41. doi:10.63125/xk6kpc72
BibTeX
@article{Mahin2024,
author = {Mahin, Kynatun and Ashfaq, Saba},
title = {Machine Learning Enhanced Intrusion Detection and Anomaly Classification for MIS-Driven Enterprise Network Security},
journal = {International Journal of Scientific Interdisciplinary Research},
year = {2024},
volume = {14},
number = {09},
pages = {01--41},
doi = {10.63125/xk6kpc72}
}
EUVIMED