A Light Gradient Boosted Model for Network Intrusion Detection

This paper investigates the development of a robust Network Intrusion Detection System (NIDS) utilizing the Light- GBM classification algorithm within the domain of data mining techniques. Extensive data preprocessing, feature engineering, and selection methods are applied to extract meaningful insights from a comprehensive dataset of web traffic attributes. Light-GBM, renowned for its meticulous approach compared to alter- native models, is specifically selected for its capacity to enhance NIDS accuracy and efficiency in discerning normal network traffic from potential intrusions. Evaluation results demonstrate the superior performance of LightGBM across diverse metrics, exhibiting high accuracy, precision, and recall. These results underscore the effectiveness of LightGBM in network intrusion detection and stress the significance of deliberate model selection for achieving high detection accuracy in practical scenarios.