Traffic Control and Accuracy Improvement of Intrusion Detection in Network using XGBoost Algorithm in Comparison with Gradient Boosting Algorithm

Aim: The study aims regulate network traffic and enhance the accuracy of intrusion detection using the XG Boost algorithm compared to the Gradient Boosting algorithm with a range of 50 to 950 estimators. Materials and Methods: This research was divided into two groups. Group 1 used the Gradient Boosting algorithm with 100 samples, which showed moderate accuracy and poor traffic control. Group 2 made use of the Boost algorithm using 100 samples, and in comparison, gave a better traffic control with accuracy. The experiment was compared on accuracy and processing time for both algorithms. Results: The XG Boost algorithm has shown much greater performance as its accuracy ranged between 91.15% and 99.90%, and the Gradient Boosting algorithm shows an accuracy between 85.53% and 95%. Besides, XG Boost ran faster and reached optimal accuracy with 950 estimators at a 0.00 significance level. Conclusion: It was observed that XG Boost outperformed the Gradient Boosting Algorithm in terms of accuracy and processing time, and it appears more efficient and reliable for intrusion detection in network traffic and for the control of the traffic.