Network Security Attack Classification: Leveraging Machine Learning Methods for Enhanced Detection and Defense

Abstract

The rapid growth and advancement of information exchange over the internet and mobile technologies have resulted in a significant increase in malicious network attacks. Machine learning (ML) algorithms have emerged as crucial tools in network security for accurately classifying and detecting these attacks, enabling effective defence strategies. In this paper, we employed ML methods such as logistic regression (LG), random forest (RF), decision tree (DT), k-nearest neighbours (KNN), and support vector machines (SVM) for building an intrusion detection system using the publicly available NSL-KDD dataset. Our proposed method utilised feature engineering and selection techniques to extract relevant features. We trained classification models and optimised their parameters using cross-validation and grid search techniques. The models exhibited robustness in identifying unseen attacks, enabling proactive defence strategies. In this paper, we contribute to the field of network security by showcasing the efficacy of machine learning methods, empowering organisations to enhance their defences and respond to threats promptly. Future research can explore advanced models and real-time monitoring techniques to develop dynamic defence mechanisms.