ABSTRACT

Network intrusion detection systems (NIDS) are critical tools for preventing malicious assaults on computer networks. Traditional NIDS, on the other hand, frequently suffer from low accuracy, large false positive rates, and poor scalability. Ensemble machine learning approaches, which combine many classifiers or models, are a promising strategy for improving the performance of NIDS. The goal of this study is to apply various ensemble machine learning algorithms to NIDS, including as bagging, boosting, stacking, and voting, and compare their performance in identifying different forms of network assaults. The NSL-KDD dataset, which contains normal and attack traffic data from a simulated network environment, will be used. The ensemble approaches will be evaluated using measures such as accuracy, precision, recall, F1-score, and ROC curve.It is expected that ensemble techniques will outperform single classifiers or models in terms of accuracy and resilience, as well as provide insights into the ideal ensemble method configuration for NIDS.