Optimizing Cybersecurity: Leveraging Support Vector Machines for Real-Time Threat Detection
Ankit Kumar Dubey, Rohit Kumar Dubey, Anil Shukla, Poornanand Dubey · 2024
In the current digital era, that is marked by the enhanced use of information technology, there are also enhanced threats which are in the form of cybercrimes. This research deals with enhancing the cyber security through Support Vector Machines (SVMs) in threat detection in real time. SVMs are a class of machine learning algorithms favored for the accuracy and efficiency they provide when it comes to the classification processes; they would be very useful in fast and accurate finding of potential threats. This forms the basis of the study under which the authors propose a new approach using SVMs in a real time data mining system that can quickly detect anomalous activities across networks. Using the detection rates for different further commonly used machine learning models, we show the advantages of our concept to the development of SVMs in view of detection accuracy, velocity and resource demand. The findings demonstrate that the use of SVM based system can alleviate the false positive incident and improve the security situation of organizations. This research extends the field of cybersecurity by developing a customizable and transferable model that can be deployed in different contexts to strengthen existing defenses against the emerging and evolving cyber threats.