Machine Learning-Based Anomaly Detection Model for Cybersecurity Threat Detection

The proliferation of cybersecurity threats continues to challenge the resilience of information systems worldwide. An effective defense against such threats requires
advanced detection methods that can predict and classify the severity of vulnerabilities with high precision. This paper proposes a sophisticated anomaly detection framework using a machine learning algorithm, aimed at identifying and categorizing cybersecurity vulnerabilities from the CISA Known Exploited Vulnerabilities catalog for 2022. The
proposed model underwent a rigorous process of preprocessing and data cleaning to ensure the integrity and suitability of the data for machine learning analysis. It has demonstrated exceptional proficiency, achieving an accuracy rate of 0.9810, alongside high precision and
recall values across various severity levels of vulnerabilities. The model's performance highlights its utility in enhancing cybersecurity measures. Therefore, the significance of this
model lies in its potential to transform the field of cybersecurity, offering a scalable, efficient tool for proactive threat detection and contributing to the fortification of
information systems against a broad spectrum of cyber threats.