摘要
针对训练集中出现未知网络应用样本的识别问题,提出一种基于改进的直推式支持向量机的未知网络应用识别算法,引入增类损失函数刻画在训练过程中新增的未知应用样本的损失代价,建立TSVM的优化问题并推导其求解过程,使得构造的分类模型能够实现对未知类别样本的识别。通过实际网络数据集进行仿真分析,结果表明所提出的算法在识别未知网络应用的可行性和有效性方面均有良好表现。
An unknown network protocol classification method based on improved transductive support vector machine learning is proposed to solve the problem of classifying augmented class when unknown network protocol data appeared in the training process.This method uses the large number of unlabeled samples to assist training classification model, where the augment loss of new un-known class samples is described by the loss augment function. TSVM( Transductive Support Vector Machine) optimization model is established and its solving process is deduced, so the decision boundary can classify the unknown class samples. The performance of the proposed method is examined in simulations with real network data sets. The experimental results illustrate the feasibility and effectiveness of the unknown network applications classified by this proposed method.
出处
《电子技术应用》
北大核心
2016年第9期95-98,共4页
Application of Electronic Technique
基金
国家自然科学基金(61309007)
国家安全重大基础研究项目(613148)
关键词
支持向量机
直推式学习
未知网络应用
流量识别
support vector machine
transductive learning
unknown network protocol
traffic classification