本文選題：流量識別 + 機(jī)器學(xué)習(xí)�。� 參考：《西安電子科技大學(xué)》2014年碩士論文

【摘要】：網(wǎng)絡(luò)應(yīng)用爆發(fā)式增長，網(wǎng)絡(luò)流量急速膨脹，大量涌現(xiàn)的新型應(yīng)用比傳統(tǒng)應(yīng)用具有更復(fù)雜的結(jié)構(gòu)和流量模式基于流量識別技術(shù)，能夠細(xì)粒度的管理和優(yōu)化網(wǎng)絡(luò)，引起了廣泛的關(guān)注其中，基于流量特征采用機(jī)器學(xué)習(xí)的流量識別技術(shù)，具有較高的準(zhǔn)確率，成為了近年來流量識別領(lǐng)域的研究熱點(diǎn) 特征選擇通過去除無關(guān)冗余的特征，獲得最優(yōu)的特征子集，基于該特征子集能夠降低學(xué)習(xí)算法的復(fù)雜度，提升分類的準(zhǔn)確率及速度 本文首先介紹了流量識別技術(shù)機(jī)器學(xué)習(xí)技術(shù)及特征選擇算法的相關(guān)概念，并簡單介紹了使用互信息進(jìn)行度量及SU算法，在此之上提出了兩種新的基于互信息的特征選擇法： 1.基于互信息的Filter式特征選擇法運(yùn)用改進(jìn)的SU算法去掉不相關(guān)的特征，并基于互信息去掉冗余特征，通過反復(fù)調(diào)整閾值進(jìn)行迭代，以提高分類準(zhǔn)確率 2.基于互信息的Wrapper式特征選擇法運(yùn)用改進(jìn)的SU算法去掉不相關(guān)的特征，并基于互信息去掉冗余特征，，直接使用分類器的分類準(zhǔn)確率作為判斷標(biāo)準(zhǔn)來指導(dǎo)算法進(jìn)行迭代，以獲得最佳閾值從而達(dá)到最好的分類效果 在UCI數(shù)據(jù)集上的實(shí)驗(yàn)結(jié)果顯示出，本文給出的兩種特征選擇算法具備較好的分類性能將本文所提出的特征選擇法應(yīng)用于網(wǎng)絡(luò)流量的類別識別中，在Andrew W.Moore數(shù)據(jù)集上的實(shí)驗(yàn)結(jié)果表明，算法在保證了分類準(zhǔn)確率的同時(shí)，取得了顯著的特征約減效果本文選出的流量識別的最優(yōu)特征子集，能夠保證較高的分類性能并大大縮短分類器的分類時(shí)間，因此為合理且有效的特征子集
[Abstract]:The network application explodes, the network traffic expands rapidly, a large number of new applications have more complex structure and traffic pattern than the traditional application, which can manage and optimize the network with fine granularity. Among them, traffic recognition technology based on machine learning, which has high accuracy rate, has become the research hotspot in the field of traffic identification in recent years. Feature selection can reduce the complexity of learning algorithm and improve the accuracy and speed of classification by removing redundant features and obtaining the optimal feature subset. In this paper, we first introduce the related concepts of machine learning and feature selection algorithm of traffic recognition technology, and simply introduce the use of mutual information to measure and Su algorithm, and then propose two new feature selection methods based on mutual information. 1. The Filter feature selection method based on mutual information uses the improved Su algorithm to remove irrelevant features, and based on mutual information to remove redundant features, iterates by adjusting the threshold value repeatedly, in order to improve the classification accuracy. 2. The Wrapper feature selection method based on mutual information uses the improved Su algorithm to remove irrelevant features, and based on mutual information to remove redundant features, the classification accuracy of classifier is directly used as the judgement standard to guide the algorithm to iterate. To get the best threshold to achieve the best classification effect. The experimental results on the UCI dataset show that the two feature selection algorithms presented in this paper have good classification performance. The proposed feature selection method is applied to the classification of network traffic. The experimental results on the Andrew W.Moore dataset show that the algorithm not only ensures the classification accuracy, but also achieves the significant feature reduction effect of the optimal feature subset selected in this paper. It can guarantee high classification performance and greatly shorten the classifier's classification time, so it is a reasonable and effective feature subset.
【學(xué)位授予單位】：西安電子科技大學(xué)
【學(xué)位級別】：碩士
【學(xué)位授予年份】：2014
【分類號】：TP393.06;TP181

【共引文獻(xiàn)】

相關(guān)期刊論文 前10條

1 周亞建;薛超;平源;;基于端口特征的P2P應(yīng)用識別方案[J];北京工業(yè)大學(xué)學(xué)報(bào);2013年11期

2 李為民;劉曉楠;繆晨;陳陸穎;雷振明;;典型業(yè)務(wù)的包長分布規(guī)律[J];電子科技大學(xué)學(xué)報(bào);2014年02期

3 錢亞冠;張e

本文編號：1951235