Network Traffic Classification Model Using Ensemble Ma-chine Learning

Abstract

Network Traffic Classification (NTC) is concerned with the identification of different types of application traffic by the examination of the data packets which are received in a communication network. This process is crucial for network management and it is getting more and more important in the last few years. The typical workflow for traffic classification includes data input, data preprocessing, attribute extraction, classification and performance analysis. It is becoming more and more rewarding the role of machine learning techniques in classifying network traffic as they become more and more advanced. This paper presents a novel paradigm intended to increase the efficacy of traffic classification. The suggested model uses a hybrid classification approach along with SMOTE to address class imbalance and Principal Component Analysis (PCA) for feature reduction. This work leverages Python, specifically when it operates in the Anaconda environment, to evaluate the efficacy of the proposed framework.