Comprehensive Analysis of Network Traffic Data

With the large volume of network traffic flow, it is necessary to preprocess raw data before classification to gain the accurate results speedily. Feature selection is an essential approach in preprocessing phase. The Principal Component Analysis (PCA) is recognized as an effective and efficient method. In this paper, we classify network traffic by using the PCA technique together with six machine learning algorithms – Naive Bayes, Decision Tree, 1-Nearest Neighbor (NN), Random Forest, Support Vector Machine (SVM) and H2O. We analyze the impact of PCA through classifying the data set by each algorithm with and without PCA. Experiments are set out by varying the size of input data sets, and the performances are measured from two metrics including overall accuracy and F-measure. The computational time is also considered in analysis phase. Our results show that Random Forest and NN are the top two algorithms among the six. Specifically, both of them behave well in classification under the most cases of input sets regardless of applying PCA. Lastly, PCA significantly boosts NN algorithms in terms of classification accuracy and shortens the classification time for Random Forest.