Proactive Cybersecurity Using Machine Learning: Threat Detection, Diagnosis, and Propagation Analysis

Deeksha Karthik, Anand Zutshi

Abstract: Cyber threats have been rising at an increasing pace, and there is a growth in demand for their detection, classification and mitigation. The following study employs machine learning technology and a data-driven approach to analyse and combat cyber threats using a dataset which comprises network traffic data, textual context and entity relationships. We performed exploratory data analysis (EDA) to identify threat patterns, their characteristics, and explore mitigation strategies to combat digital threats. The threats have been classified based on textual data to predict diagnoses and solutions. Furthermore, network and relationship analysis have been used to identify the communication pathways, entities, and understand the propagation of cyber threats visually using graphs. Our results show that combining text analysis, graph models, and machine learning improves the accuracy of cyber threat detection, diagnosis, and mitigation. This research contributes to the field of cybersecurity by offering a framework to strengthen proactive defences against evolving cyber threats.

Keywords: Cybersecurity, Machine Learning, Threat Classification, Graph-Based Analysis, Network Propagation

How to Cite?: Deeksha Karthik, Anand Zutshi, "Proactive Cybersecurity Using Machine Learning: Threat Detection, Diagnosis, and Propagation Analysis", Volume 14 Issue 9, September 2025, International Journal of Science and Research (IJSR), Pages: 557-565, https://www.ijsr.net/getabstract.php?paperid=SR25816183803, DOI: https://dx.doi.org/10.21275/SR25816183803