Comparative Analysis of Machine Learning Models for Network Traffic Forecasting using Chaos Engineering and Beta Testing

This study presents a comparative analysis of machine learning models, including Neural Prophet, XGBoost, Random Forest Regressor, and LSTM, for network traffic forecasting. It explores the application of chaos engineering principles and beta testing on these models, with the goal of optimizing network performance and ensuring reliable service delivery. Accurate network traffic forecasting is crucial for maintaining efficient operations. The models are evaluated using Mean Absolute Error (MAE) and Random Tree Order (RTO) as performance metrics. The results highlight which models demonstrate superior performance under chaos-induced scenarios, such as traffic surges and drops, in terms of robustness and predictive accuracy. This work enhances our understanding of machine learning model efficiency in dynamic and unpredictable networking conditions and outlines implications for network optimization.