ABSTRACT
The network is analyzed manually or some fixed abnormal patterns are given in the system in case of traditional model such that an attack present in the system can be identified. Lately, the traffic on network is improved by the use of internet and the threats of attacking are identified since the policy can be accessed easily. Thus, the network analyst can improve the data through these activities and detecting the intrusions becomes difficult. The network traffic classification has various phases in which dataset can be proposed for the elimination of missing and redundant values from the dataset. The Support Vector Machine classier is applied in the existing methodology for the network traffic classification. The voting classification is implemented in this research work for the network traffic classification. The performance is analyzed with regard to accuracy, precision and recall. It is analyzed that voting classification gives high performance as compared to SVM classifier.
