Advanced Search
Volume 32 Issue 6
Jun.  2010
Turn off MathJax
Article Contents
Tian Jiang, Gu Hong. Outlier One Class Support Vector Machines[J]. Journal of Electronics & Information Technology, 2010, 32(6): 1284-1288. doi: 10.3724/SP.J.1146.2009.00861
Citation: Tian Jiang, Gu Hong. Outlier One Class Support Vector Machines[J]. Journal of Electronics & Information Technology, 2010, 32(6): 1284-1288. doi: 10.3724/SP.J.1146.2009.00861

Outlier One Class Support Vector Machines

doi: 10.3724/SP.J.1146.2009.00861
  • Received Date: 2009-06-09
  • Rev Recd Date: 2009-10-16
  • Publish Date: 2010-06-19
  • One-Class Support Vector Machines (OCSVMs) distinguish outliers by computing a hyper-plane in feature space. The choice of the origin as separation point is arbitrary, which affects the decision boundary, and the distribution of samples has impact on the performance. Expanding the algorithm into solving two-class classification problems overcomes the drawbacks to a certain degree. However, the class imbalance problem is serious and the labeled outliers are rare or even non-existing. In this paper, a new Outlier OCSVM is proposed and a framework is designed for unsupervised outlier detection. Respectively scored by distance from hyper-plane and probabilistic output value, two definitions of outlier degree are presented. After picking out some suspicious outliers via combining the two criterions of outlier degree, the adjusted Outlier OCSVM starts the training operations, two parts of the dataset are updated interactively through comparison of the outputs. Experiment results on benchmark datasets show that the method can effectively improve the detection rate and reduce false positive rate, easy and reliable.
  • loading
  • Han J and Kamber M. Data Mining: Concepts and Techniques[M]. San Francisco: Morgan Kaufmann Publishers, 2006: 451-458.[2]倪巍伟, 陈耿, 陆介平, 吴英杰, 孙志挥. 基于局部信息熵的加权子空间孤立点检测算法[J].计算机研究与发展, 2008, 45(7): 1189-1194.Ni W W, Chen G, Lu J P, Wu Y J, and Sun Z H. Local entropy based weighted subspace outlier mining algorithm[J]. Journal of Computer Research and Development, 2008, 45(7): 1189-1194.[3]薛安荣, 鞠时光, 何伟华, 陈伟鹤. 局部孤立点挖掘算法研究[J]. 计算机学报, 2007, 30(8): 1455-1463.Xue A R, Ju S G, He W H, and Chen W H. Study on algorithms for local outlier detection[J]. Chinese Journal of Computers, 2007, 30(8): 1455-1463.[4]庞彦伟, 刘政凯. 一种自动抑制孤立点的子空间学习方法[J].电子与信息学报.2008, 30(1):176-179浏览Pang Y W and Liu Z K. Automatically outlier-resisting subspace learning[J]. Journal of Electronics and Information Technology, 2008, 30(1): 176-179.[5]Giacinto G, Perdisci R, Del Rio M, and Roli F. Intrusion detection in computer networks by a modular ensemble of one-class classifiers[J].Information Fusion.2008, 9(1):69-82[6]Chandola V, Banerjee A, and Kumar V. Anomaly detection: A survey[J]. ACM Computing Survey, 2009, 41(3): 1-58.[7]Scholkopf B, Williamson R C, Smola A J, Shawe-Taylor J, and Platt J. Support vector method for novelty detection[J]. Advances in Neural Information Processing Systems,2000,12(3): 582-588.[8]Scholkopf B, Platt J C, Shawe-Taylor J, Smola A J, and Williamson R C. Estimating the support of a high-dimensional distribution[J].Neural Computation.2001, 13(7):1443-1471[9]Eskin E, Arnold A, Prerau M, Portnoy L, and Stolfo S. Applications of Data Mining in Computer Security[M]. Norwell Massachusetts:Kluwer Academic Publishers, 2002: 77-87.[10]Manevitz L M and Yousef M. One-class SVMs for document classification[J].Journal of Machine Learning Research.2002, 2(2):139-154[11]He J R, Li M J, Li Z W, Zhang H J, Tong H H, and Zhang C S. Pseudo relevance feedback based on iterative probabilistic one-class SVMs in web image retrieval, In: Advances in Multimedia Information Processing - Pcm 2004, Pt 2, Proceedings[C]. Lecture Notes in Computer Science, 2004, 3332: 213-220.[12]Scholkopf B, Platt J, and Smola A J. Kernel method for percentile feature extraction[R]. Microsoft Research Ltd, 2000.[13]Munoz A and Moguerza J M. Estimation of high-density regions using one-class neighbor machines[J].IEEE Transactions on Pattern Analysis and Machine Intelligence.2006, 28(3):476-480
  • 加载中

Catalog

    通讯作者: 陈斌, bchen63@163.com
    • 1. 

      沈阳化工大学材料科学与工程学院 沈阳 110142

    1. 本站搜索
    2. 百度学术搜索
    3. 万方数据库搜索
    4. CNKI搜索

    Article Metrics

    Article views (3579) PDF downloads(1683) Cited by()
    Proportional views
    Related

    /

    DownLoad:  Full-Size Img  PowerPoint
    Return
    Return