Hyper Parameter Optimization for Ensemble Techniques in Classifying QUIC Traffic

The number of Internet-based devices is increasing exponentially in day-to-day life. Higher levels of usage results in increased network traffic. It affects the network performance rate and quality, thereby leaving a threat to security. Network traffic classification is the process that classifies different types of network traffic and identifies solutions to overcome these issues. Efficient classification of Quick UDP Internet Connections (QUIC) traffic is crucial for network management. This study focuses on optimizing classification accuracy through ensemble techniques and hyperparameter tuning. Ensemble methods, like bagging and boosting, combine multiple base models to enhance the performance. Hyperparameter optimization fine-tunes these models, controlling their behavior and improving classification accuracy. The research contributes insights into effective ensemble modeling for network protocols, offering practical guidance for QUIC traffic management and optimization. It underscores the significance of hyperparameter optimization in advancing QUIC network classification accuracy, which is crucial for modern internet communication.