一种基于特征集构建的Bagging集成方法及其在流量分类中的应用

doi:10.11959/j.issn.1000-0801.2018094

Abstract

Abstract:

Bagging is a classic ensemble approach,whose effectiveness depends on the diversity of component base classifiers.In order to gain the largest diversity,employing genetic algorithms to get independent feature subset for each base classifier was proposed.Meanwhile,for better generalization,the optimal weights for the base classifiers according to their predictive performance were selected.Finally,refined Bagging ensemble based on simple Softmax regression was applied successfully in traffic classification.The experiment result shows that the proposed approach can get more improvement than the original Bagging ensemble in classification performance,and is better than the random-forests to a certain extent.

Key words: Bagging ensemble, feature subset, genetic algorithm, traffic classification

CLC Number:

TP301.6

Yaguan QIAN,Xiaohui GUAN,Shuhui WU,Bensheng YUN,Dongxiao REN. An approach of Bagging ensemble based on feature set and application for traffic classification[J]. Telecommunications Science, 2018, 34(4): 41-48.

Figures/Tables 5

References 14

[1]	ZHOU Z H . Ensemble methods:foundations and algorithms[M]. Boca Raton: CRC PressPress, 2012.
[2]	LI H , WANG X , DING S . Research and development of neural network ensembles:a survey[J]. Artificial Intelligence Review, 2017: 1-25.
[3]	AMOZEGAR M , KHORASANI K . An ensemble of dynamic neural network identifiers for fault detection and isolation of gas turbine engines[J]. Neural Networks, 2016(76): 106-121.
[4]	INOUE H . Fast and accurate inference with adaptive ensemble prediction in image classification with deep neural networks[J]. arXiv preprint arXiv:1702.08259, 2017.
[5]	WANG Q , LUO Z H , HUANG J C ,et al. A novel ensemble method for imbalanced data learning:bagging of extrapolation-SMOTE SVM[J]. Computational Intelligence and Neuroscience, 2017(3):1827016.
[6]	高文, 钱亚冠, 吴春明 ,等. 网络流量特征选择方法中的分治投票策略研究[J]. 电子学报, 2015,43(4): 795-799.
	GAO W , QIAN Y G , WU C M ,et al. The divide-conquer and voting strategy for traffic feature selection[J]. Acta Electronica Sinica, 2015,43(4): 795-799.
[7]	钱亚冠, 张旻 . 基于过抽样技术的 P2P 流量识别方法[J]. 电信科学, 2014,30(4): 109-113.
	QIAN Y G , ZHANG M . P2P traffic identification based over-sampling technique[J]. Telecommunications Science, 2014,30(4): 109-113.
[8]	KEARNS M . Learning Boolean formulae or finite automata is as hard as factoring[R]. Technical Report TR-14-88 Harvard University Aiken Computation Laboratory, 1988.
[9]	KEARNS M , VALIANT L . Cryptographic limitations on learning Boolean formulae and finite automata[J]. Journal of the ACM (JACM), 1994,41(1): 67-95.
[10]	EFRON B , TIBSHIRANI R . An introduction to the bootstrap[M]. New York: Chapman ＆ HallPress, 1993.
[11]	TONGAONKAR A , TORRES R , ILIOFOTOU M ,et al. Towards self-adaptive network traffic classification[J]. Computer Communications, 2015(56): 35-46.
[12]	SOYSALA M , SCHMIDT E G . Machine learning algorithms for accurate flow-based network traffic classification:evaluation and comparison[J]. Performance Evaluation, 2010,67(6): 451-467.
[13]	SINGH H , . Performance analysis of unsupervised machine learning techniques for network traffic classification[C]// 2015 Fifth International Conference on Advanced Computing＆Communication Technologies (ACCT),May 15-16,2015,Haryana,India. Piscataway:IEEE Press, 2015: 401-404.
[14]	MOORE A W . Dataset[EB]. 2017

Metrics

Recommended 0

No Suggested Reading articles found!

	召回率		精度		F-measure
	Bagging	Bagging+	Bagging	Bagging+	Bagging	Bagging+
WWW	99.1%	99.9%	97.6%	98.4%	98.3%	99.1%
mail	97.8%	100.0%	97.4%	99.6%	97.6%	99.8%
FTP-control	100.0%	100.0%	98.0%	98.0%	99.0%	99.0%
FTP-PASV	15.7%	60.5%	76.5%	84.4%	26.5%	67.5%
P2P	19.2%	51.1%	50.4%	98.0%	27.8%	67.1%
database	100.0%	100.0%	82.9%	88.9%	90.7%	94.1%
FTP-data	96.7%	100.0%	97.5%	99.8%	97.1%	99.9%
multimedia	33.3%	94.7%	49.2%	99.3%	39.7%	96.9%

	召回率		精度		F-measure
	随机森林	Bagging+	随机森林	Bagging+	随机森林	Bagging+
WWW	97.1%	99.9%	99.3%	98.4%	98.2%	99.1%
mail	100.0%	100.0%	98.7%	99.6%	99.3%	99.8%
FTP-control	96.0%	100.0%	100.0%	98.0%	98.0%	99.0%
FTP-PASV	68.8%	60.5%	82.7%	84.4%	70.8%	67.5%
P2P	39.4%	51.1%	77.1%	98.0%	52.1%	67.1%
database	98.5%	100.0%	100.0%	88.9%	99.2%	94.1%
FTP-data	100.0%	100.0%	70.9%	99.8%	83.0%	99.9%
multimedia	92.0%	94.7%	91.4%	99.3%	91.7%	96.9%

[1]	Lu LIU, Dan YANG, Ruijie CHEN, Jia LI, Xi ZHOU. Research on POI quality prediction based on KPCA-GA-BP neural network [J]. Telecommunications Science, 2023, 39(1): 108-116.
[2]	Kun MA, Lingyu XU, Xiaoping SHEN, Zhicheng GONG, Jianping LAN, Shuangxi CHEN, Jun QIAN. Virtual machine scheduling model based on Shapley value modified genetic algorithm in cloud computing [J]. Telecommunications Science, 2022, 38(12): 1-10.
[3]	Peng ZHANG, Xiaoping JIN, Dongxiao CHEN. A genetic algorithm based method of optimizing dispersion matrix for RDSM system [J]. Telecommunications Science, 2022, 38(12): 46-55.
[4]	Xiaoping JIN, Jiayu LIU, Chen JIANG, Qiang GUO. Joint optimization of dispersion matrix and 3D constellation for STSK system based on improved genetic algorithm [J]. Telecommunications Science, 2021, 37(9): 86-94.
[5]	Yue GU, Dan LI, Kaihui GAO. Research on network traffic classification based on machine learning and deep learning [J]. Telecommunications Science, 2021, 37(3): 105-113.
[6]	Xiaohui YAO,Qing LI,Kunkun SUN. A domain adaptive hybrid genetic algorithm and its verification in security service function chain orchestration [J]. Telecommunications Science, 2020, 36(5): 16-24.
[7]	Wenyu HU,Wenju WANG. Empirical research on the diffusion effect of ICT innovation diffusion and prediction in China [J]. Telecommunications Science, 2020, 36(1): 66-76.
[8]	Kai ZHOU. Nodes deployment strategy for underwater wireless sensors networks based on grids [J]. Telecommunications Science, 2018, 34(11): 48-58.
[9]	Chang CAI,Yafang WANG,Bingmei MIAO,Hui JIANG. Dynamic spectrum allocation for cognitive radio sensor networks based on improved genetic algorithm [J]. Telecommunications Science, 2017, 33(8): 85-93.
[10]	Heng ZHOU,Zhixian CHANG,Wujun YANG,Juan GUO. An orchestration algorithm of 5G network slicing [J]. Telecommunications Science, 2017, 33(8): 130-137.
[11]	Guoling ZHANG. An emotional neural network based approach for wind power prediction [J]. Telecommunications Science, 2017, 33(3): 168-172.
[12]	Xiaoan BAO,Xue WEI,Lei CHEN,Guoheng HU,Na ZHANG. Load balancing method of service cluster based on mean-variance [J]. Telecommunications Science, 2017, 33(1): 1-8.
[13]	Qing SHEN,Yunliang JIANG,Zhangguo SHEN,Jungang LOU. Test case minimizing based on combination chaos genetic algorithm [J]. Telecommunications Science, 2016, 32(6): 93-102.
[14]	Zhen LIU,Ruoyu WANG. Internet traffic classification method based on behavior feature learning [J]. Telecommunications Science, 2016, 32(6): 143-152.
[15]	Yaguan QIAN,Xiaohui GUAN,Bensheng YUN,Qiong LOU,Pengfei MA. Internet traffic classification using SVM with flexible feature space [J]. Telecommunications Science, 2016, 32(5): 105-113.

An approach of Bagging ensemble based on feature set and application for traffic classification

RichHTML

PDF下载

Knowledge

Abstract

Cite this article

share this article

Figures/Tables 5

References 14

Related Articles 15

Metrics

Recommended 0