Network Anomaly Detection Using Borderline SMOTE Algorithm and Support Vector Machines

Network Security is a major challenge in the digital world. Intrusion is common in many applications and intruders are sophisticated enough to change their attack pattern very often. To address this issue, the development of a model for the detection of network anomalies and intrusions is required. The proposed approach utilizes the Borderline Synthetic Minority Over-Sampling Technique (SMOTE) in augmentation with Support Vector Machines (SVM) to enhance anomaly detection capabilities. This is achieved by intelligently oversampling the minority class using SMOTE and training SVM, the proposed model exhibits a robust defence mechanism against network intruders. The utilization of these advanced techniques aims to increase the accuracy and efficiency of anomaly detection, minimizing false positives and ensuring prompt response to genuine threats. The study produces results that add to the ongoing efforts to secure data in the digital age, by combining SMOTE and SVM for network intrusion detection.