摘要
提出一种新的基于神经网络集成的P2P流量识别方法,利用CFS特征选择算法提取P2P流量特征,使用动态加权集成方法将6个神经网络集成应用于P2P流量识别。通过在实际网络流数据集上与单一BP神经网络、决策树、朴素贝叶斯和支持向量机算法的对比实验,结果表明该方法具有较高的P2P流量识别准确率和稳定性。
A novel P2P traffic identification method based on neural network ensemble is proposed. A P2P flow detection model is developed by using correlation-based feature selection (CFS) algorithm to extract P2P flow characteristics, and utilizing six ensemble neural networks by dynamic weighted integration method. Through experimental comparison between this proposed model and traditional methods, such as single BP neural network, decision tree, bayesian, and support vector machine, it is shown that the proposed method has a better P2P traffic identification accuracy and stability.
出处
《南京邮电大学学报(自然科学版)》
2010年第3期79-83,共5页
Journal of Nanjing University of Posts and Telecommunications:Natural Science Edition
基金
国家自然科学基金(60973139
60773041)
国家高技术研究发展计划(863计划)(2007AA01Z404)
江苏省自然科学基金(BK2008451)
2006江苏省软件专项
省级现代服务业发展专项资金
中国博士后科学基金(0801019C)
江苏省博士后科研资助计划(20090451240
20090451241)
江苏高校科技创新计划(CX09B-153Z
CX08B-086Z)
江苏省"六大人才高峰"计划(2008118)
江苏省计算机信息处理技术重点实验室基金(2010)资助项目
关键词
神经网络
集成学习
流量识别
P2P
neural network
ensemble learning
traffic identification
P2P