Network Security Attack Detection Using Machine Learning Algorithms

This research study deals with network security attack detection using machine learning algorithms. Random Forest, Support Vector Machine, and Decision Tree algorithms are analyzed in the paper. In this paper, we have applied the NSL-KDD dataset and compared the performance of the stated algorithms to find the most accurate detection among various types of network attacks. All these three performed very well in our experiments; however, Decision Tree algorithms were always outperforming others in accuracy, precision, and recall. Excellence in performance by the Decision Tree is because a decision tree can handle complicated decision-making processes and is robust enough to manage various attack patterns. This research points out that, compared with other machine learning techniques, decision trees may enhance intrusion detection systems both in effectiveness and dependability, making the concept of defense against cyber threats more watertight.