An optimal and stable feature selection approach for traffic classification based on multi-criterion fusion

Fahad, Adil; Tari, Zahir; Khalil, Ibrahim; Almalawi, Abdulmohsen; Zomaya, Albert Y.

doi:10.1016/j.future.2013.09.015

Cited by 46 publications

(33 citation statements)

References 23 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…3) A series of experiments are carried out on real-world network traffic datasets, in order to evaluate the performance of our approach, including the classification accuracy, efficiency and robustness. We compare our approach against three techniques, BFS (balanced feature selection) [27], WSU_AUC (weighted symmetric uncertainty_area under ROC curve) [32] and GOA (global optimization approach) [15], presented recently for network traffic classification task. Classification results show that COFS indeed improves our previous work BFS.…”

Section: ) a New Algorithm Called Class-oriented Feature Selection (mentioning

confidence: 99%

“…Moore et al [11] used FCBF (fast correlation-based filter) to remove the redundant and irrelevant features, and then evaluated the classification accuracy of Naïve Bayes for searching final optimal features. Recently, Fahad et al [15] proposed a new approach to address the problem that feature selection techniques were often sensitive to a small variation in the traffic data. Five well-known feature selection algorithms were respectively used to search a feature subset, and an initial feature subset was built through selecting high frequency features from the five feature subsets.…”

Section: Related Work On Feature Selectionmentioning

confidence: 99%

“…In the first part (1-4), it calculates the WSU value of each feature on NS, and inserts each feature into list in descending order according to their WSU values. In the second part (5)(6)(7)(8)(9)(10)(11)(12)(13)(14)(15)(16)(17)(18)(19), it further processes the ordered list to remove redundant features and only keeps predominant ones among all the selected features. A feature f p that has been determined to be a predominant feature can always be used to filter out other features that are ranked lower than f p .…”

Section: Algorithm 1 Feature Selection For Each Classmentioning

confidence: 99%

“…A variety of features have been extracted from traffic flows in recent years. However, the presence of irrelevant or redundant features degrades the classification accuracy, maximizes the training and testing processing time of the classifier, and finally increases its storage requirements [15]. With the object of searching a subset of good features, feature selection techniques can play an effective role in removing irrelevant and redundant features.…”

Section: Introductionmentioning

confidence: 99%

See 3 more Smart Citations

A class-oriented feature selection approach for multi-class imbalanced network traffic datasets based on local and global metrics fusion

et al. 2015

View full text Add to dashboard Cite

Section: ) a New Algorithm Called Class-oriented Feature Selection (mentioning

confidence: 99%

Section: Related Work On Feature Selectionmentioning

confidence: 99%

Section: Algorithm 1 Feature Selection For Each Classmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

A class-oriented feature selection approach for multi-class imbalanced network traffic datasets based on local and global metrics fusion

et al. 2015

View full text Add to dashboard Cite

“…However, the filter techniques eliminate both irrelevant and redundant attributes from a local perspective, and thus it can be tricked in a situation where the dependence between a pair of attributes is weak, but the total inter-correlation of one attribute to the others is strong. Thus, in this paper, we introduce a new FS approach to select informative attributes from a global perspective [5], [11]. The process of discarding irrelevant and redundant attributes from a global perspective and only keeping the optimal attributes is presented in Table I.…”

Section: The Proposed Hybrid Clustering-classification Approachmentioning

confidence: 99%

CluClas: Hybrid clustering-classification approach for accurate and efficient network classification

Fahad

Alharthi

Tari

et al. 2014

39th Annual IEEE Conference on Local Computer Networks

Self Cite

View full text Add to dashboard Cite

The traffic classification is the foundation for many network activities, such as Quality of Service (QoS), security monitoring, Lawful Interception and Intrusion Detection Systems (IDS). A recent statistics-based approach to address the unsatisfactory results of traditional port-based and payloadbased approaches has attracted attention. However, the presence of non-informative attributes and noise instances degrade the performance of this approach. Thus, to address this problem, in this paper, we propose a hybrid clustering-classification approach (called CluClas) to improve the accuracy and efficiency of network traffic classification by selecting informative attributes and representative instances. An extensive empirical study on four traffic data sets shows the effectiveness of our proposed approach.

show abstract

An accurate traffic classification model based on support vector machines

Cao

Fang

et al. 2017

Int. J. Network Mgmt

View full text Add to dashboard Cite

SummaryNetwork traffic classification is a fundamental research topic on high-performance network protocol design and network operation management. Compared with other state-of-the-art studies done on the network traffic classification, machine learning (ML) methods are more flexible and intelligent, which can automatically search for and describe useful structural patterns in a supplied traffic dataset. As a typical ML method, support vector machines (SVMs) based on statistical theory has high classification accuracy and stability. However, the performance of SVM classifier can be severely affected by the data scale, feature dimension, and parameters of the classifier. In this paper, a real-time accurate SVM training model named SPP-SVM is proposed. An SPP-SVM is deducted from the scaling dataset and employs principal component analysis (PCA) to extract data features and verify its relevant traffic features obtained from PCA. By employing PCA algorithm to do the dimension extraction, SPP-SVM confirms the critical component features, reduces the redundancy among them, and lowers the original feature dimension so as to reduce the over fitting and increase its generalization effectively. The optimal working parameters of kernel function used in SPP-SVM are derived automatically from improved particle swarm optimization algorithm, which will optimize the global solution and make its inertia weight coefficient adaptive without searching for the parameters in a wide range, traversing all the parameter points in the grid and adjusting steps gradually. The performance of its two-and multi-class classifiers is proved over 2 sets of traffic traces, coming from different topological points on the Internet. Experiments show that the SPP-SVM's two-and multi-class classifiers are superior to the typical supervised ML algorithms and performs significantly better than traditional SVM in classification accuracy, dimension, and elapsed time.

show abstract

An optimal and stable feature selection approach for traffic classification based on multi-criterion fusion

Cited by 46 publications

References 23 publications

A class-oriented feature selection approach for multi-class imbalanced network traffic datasets based on local and global metrics fusion

A class-oriented feature selection approach for multi-class imbalanced network traffic datasets based on local and global metrics fusion

CluClas: Hybrid clustering-classification approach for accurate and efficient network classification

An accurate traffic classification model based on support vector machines

Contact Info

Product

Resources

About