Performance Analysis of Semi-supervised Intrusion Detection System

Supervised learning algorithm for Intrusion Detection needs labeled data for training. Lots of data is available through internet, network and host. But this data is unlabeled data. The availability of labeled data needs human expertise which is costly. This is the main hurdle for developing supervised intrusion detection systems. We can intelligently use both labeled and unlabeled data for intrusion detection. Semisupervised learning has attracted the attention of the researcher working in Intrusion Detection using machine learning. Our goal is to improve the classification accuracy of any given supervised classifier algorithm by using the limited labeled data and large unlabeled data. The key advantage of the proposed semisupervised learning approach is to improve the performance of supervised classifier. The results show that the performance of the proposed semi-supervised algorithm is better than the stateof theart supervised learning algorithms. We compare the performance of our DS-AdaBoost algorithm as well as 5 standard algorithms available in WEKA for supervised and semi-supervised approach.

[1]  Avrim Blum,et al.  The Bottleneck , 2021, Monopsony Capitalism.

[2]  Yuh-Jye Lee,et al.  Semi-supervised Learning for False Alarm Reduction , 2010, ICDM.

[3]  Wang Min,et al.  Applying Semi-supervised Cluster Algorithm for Anomaly Detection , 2010, 2010 Third International Symposium on Information Processing.

[4]  S.F. Yusufovna,et al.  Integrating Intrusion Detection System and Data Mining , 2008, 2008 International Symposium on Ubiquitous Multimedia Computing.

[5]  Andrew H. Sung,et al.  Feature Selection for Intrusion Detection with Neural Networks and Support Vector Machines , 2003 .

[6]  David Yarowsky,et al.  Unsupervised Word Sense Disambiguation Rivaling Supervised Methods , 1995, ACL.

[7]  A. Karr,et al.  Computer Intrusion: Detecting Masquerades , 2001 .

[8]  Xiaojin Zhu,et al.  --1 CONTENTS , 2006 .

[9]  Tsuhan Chen,et al.  Semi-supervised co-training and active learning based approach for multi-view intrusion detection , 2009, SAC '09.

[10]  Yi Liu,et al.  SemiBoost: Boosting for Semi-Supervised Learning , 2009, IEEE Transactions on Pattern Analysis and Machine Intelligence.

[11]  Vladimir Vapnik,et al.  Statistical learning theory , 1998 .

[12]  Arthur B. Maccabe,et al.  The architecture of a network level intrusion detection system , 1990 .

[13]  Olfa Nasraoui,et al.  Anomaly detection based on unsupervised niche clustering with application to network intrusion detection , 2004, Proceedings of the 2004 Congress on Evolutionary Computation (IEEE Cat. No.04TH8753).

[14]  Zoubin Ghahramani,et al.  Combining active learning and semi-supervised learning using Gaussian fields and harmonic functions , 2003, ICML 2003.

[15]  Zoubin Ghahramani,et al.  Towards semi-supervised classification with Markov random fields , 2002 .

[16]  Thomas L. Griffiths,et al.  Semi-Supervised Learning with Trees , 2003, NIPS.

[17]  Neil D. Lawrence,et al.  Semi-supervised Learning via Gaussian Processes , 2004, NIPS.