Improved Machine Learning Models for Early Prediction of Cybersecurity Breaches in Modern Networked Systems

Abiola O. Akinyemi

doi:10.56201/rjpst.vol.8.no12.2025.pg9.15

📄 Download Paper

Improved Machine Learning Models for Early Prediction of Cybersecurity Breaches in Modern Networked Systems

Abiola O. Akinyemi

Abstract

Cybersecurity breaches are becoming more complex, frequent, and widespread as modern networked systems grow. The rise in cyberattacks demands effective early-warning systems. This study presents improved machine learning (ML) models that aim to predict breaches by spotting high-risk behaviors before an attack is fully launched. The paper combines optimized feature extraction, mixed learning structures, and flexible detection processes to boost prediction accuracy and minimize false alerts. Experimental tests using simulated datasets show that the proposed model can identify early signs of breaches with much better sensitivity. The developed framework offers a reliable and adaptable way to protect modern networked systems.

Keywords

early breach predictionmachine learningcyber securityintrusion detectiongraph learningtemporal patternscyber-threat analytics.

References

Chaudhuri, S., & Rai, P. (2022). Handling data imbalance in cybersecurity analytics using SMOTE variants. Journal of Intelligent Security Systems, 12(4), 88-97. Hashim, R., Al-Yahya, M., & Farhan, S. (2022). LSTM-based intrusion prediction for modern IoT networks. IEEE Access, 10, 77122-77135. Kumar, P., & Patel, D. (2021). Machine learning-driven intrusion detection for cloud systems. International Journal of Cyber Defense, 5(1), 33-48. Lim, H., & Roberts, T. (2023). Predictive modelling for pre-attack detection in enterprise networks. Computers & Security, 121, 103041. Musa, I., & Wang, X. (2024). Early breach detection using temporal ML classifiers in large-scale networks. Journal of Network Security Engineering, 18(1), 17-31. Oladipo, F., & Chen, Y. (2022). Hybrid cybersecurity models for predictive intrusion analysis. Information Security Review, 14(2), 65-78. Rahman, J., Singh, A., & Omar, R. (2023). Limitations of machine learning-based IDS in modern attack surfaces. International Journal of Secure Computing, 7(3), 99-115. Singh, A., Kaur, P., & Li, J. (2021). Evaluating machine learning... (and more from the snippet)

Improved Machine Learning Models for Early Prediction of Cybersecurity Breaches in Modern Networked Systems

Abstract

Keywords

References

More Articles from RESEARCH JOURNAL OF PURE SCIENCE AND TECHNOLOGY