期刊文献+
任意字段
题名或关键词
题名
关键词
文摘
作者
第一作者
机构
刊名
分类号
参考文献
作者简介
基金资助
栏目信息

一种面向间隙核函数的快速算法 被引量：1

A Fast Algorithm for Gapped Kernels
下载PDF
导出
摘要 间隙核是一种应用非常广泛的字符串核,在文本分类和蛋白质分类中都取得了很好的效果.本文提出了一种应用在入侵检测领域的间隙核,称为长度加权核.并且提出了一种基于后缀核的动态规划算法,能够有效计算变长度加权核.另外,本文提出了一种位并行算法,能够加速定长度加权核的计算.实验表明在满足位并行的条件下这种快速算法比现有的几种计算间隙核的算法更为快速,而且应用在入侵检测中能够取得较好的效果. So far the gapped kemels are used in many fields,such as text classification and protein classification.In this paper,a new kind of gapped kernel is presented, which is called length-weighted kernel, including p-length-weighted and all-length-weighted kemels. Length-weighted kernels can be used to detect intrusion process. Furthermore, a dynamic programming algorithm based on suffix kernel is proposed to compute the length-weighted kemels.Moreover, a bit-parallel technique is used to reduce the complexity of p-length-weighted kernel. The empirical results suggest that this bit-parallel technique algorithm outperforms the other approaches in some cases where the necessary condition of using bit-parallel technique can be satisfied, and that the new kernels can achieve better performance than others gapped kernels.
作者 尹传环 田盛丰 牟少敏
机构地区 北京交通大学计算机与信息技术学院
出处 《电子学报》 EI CAS CSCD 北大核心 2007年第5期875-881,共7页 Acta Electronica Sinica
基金 国家自然科学基金(No.60442002) 北京交通大学科技基金(No.2004SM010)
关键词 核方法 字符串核 间隙核 位并行 kemel methods string kemels gapped kernels bit-parallel technique
分类号 TP18 [自动化与计算机技术—控制理论与控制工程]
  • 相关文献

参考文献13

  • 1Vapnik V.The Nature of Statistical Learning Theory[M].New York:Springer Verlag,1995.
  • 2Cristianini N,Shawe-Taylor J.An introduction to Support Vector Machines[M].Cambridge,UK:Cambridge University Press,2000.
  • 3Shawe-Taylor J,Cristianini N.Kernel Methods for Pattern Analysis[M].Cambridge,UK:Cambridge University Press,2004.
  • 4Watkins C.Dynamic alignment kernels[A].In Smola A J,et al eds.,Advances in Large Margin Classifiers[C].Massachusetts:MIT Press,2000.39-50.
  • 5Haussler D.Convolution kernels on discrete structures.Technical Report UCSC-CRL-99-10[R].Department of Computer Science,University of California at Santa Cruz,Santa Cruz,CA,1999.
  • 6Joachims T.Learning to Classify Text Using Support Vector Machines:Methods,Theory and Algorithms[M].MA:Kluwer Academic Publishers,2002.
  • 7Leslie C,Kuang R.Fast String kernels using inexact matching for protein sequences[J].Journal of Machine Learning Research,2004,5(11):1435-1455.
  • 8Lodhi H,Saunders C,Shawe-Taylor J,Cristianini N,Watkins C.Text classification using string kernels[J].Journal of Machine Learning Research,2002,2(2):419-444.
  • 9Rousu J,Shawe-Taylor J.Efficient computation of gapped substring kernels on large alphabets[J].Journal of Machine Learning Research,2005,6(9):1323-1344.
  • 10Myers G.A fast bit-vector algorithm for approximate string matching based on dynamic programming[J].Journal of the ACM,1999,46(3):395-415.

二级参考文献12

  • 1S Forrest, et al. A sense af seaf for unix processes[A]. John McHugh IEEE Symposium on Security and Privacy, Proceedings[C]. Oakland CA:IEEE Computer Society Press, 1996.120 - 128.
  • 2A P Kosoresow, S A Hofmey. Intrusion detection via system call traces[J]. IEEE Software, 1997,14(5) :35 - 42.
  • 3W Lee, et al. Learning patteans from UNIX process execution traces forintrusion detection [A ]. AAAI Wodtshop on AI Approaches to Fraud Detection and Risk Management [C ]. Rhode laland: AAAI Press,1997.50 - 56.
  • 4M Asaka, et al. A new intrusion detection method based on discriminant analysis [J]. IEICE Tram. on Information & Systems, 2001, E-84-B(5) :570 - 577.
  • 5Yihua Liao, V Rao Vemuri. Using text categorization techniques for intrusion detection [ A ]. 11th USENIX Security Symposium [ C ]. San.Francisco, 2002.
  • 6H Debar, et al. Fixed vs. Variable-length pattexns for detecting suspicious process behavior [A] .5th European Symposium on Research in Computer Security [ C ]. Belgium: Springer-Verlag, 1998.1 - 15.
  • 7C Michael, A Ghosh. Two state-based approaches to program-based anomaly detection [DB/OL]. www. acsac, org/2000/parpers/96. pdf.
  • 8R Sekar, et al. A fast automaton-tin.seal method for detecting anomalous program behaviors [ A]. Roger Needham,IEEE Symposium on Securityand Privacy [C]. California: IEEK Computer Society Press,2001. 144- 155.
  • 9Aho A V, M J Corasick. Efficient suing matching: an aid to bibliographic search [J] .Communications of the ACM, 1975:333 - 340.
  • 10http://www.cs. unm. edu/- immsec/data [OL].

共引文献16

同被引文献39

  • 1常群,王晓龙,林沂蒙,王熙照,Daniel S.Yeung.支持向量分类和多宽度高斯核[J].电子学报,2007,35(3):484-487. 被引量：10
  • 2Baram Y.Learning by kernel polarization[J].Neural Com-putation,2005,17(6):1264-1275.
  • 3Nguyen C H,Ho T B.Kernel matrix evaluation[C].Hy-derabad,India:Proc of the 20th International JointConference on Artificial Intelligence,2007:987-992.
  • 4WANG T,TIAN S,HUANG H,et al.Learning by localkernel polarization[J].Neurocomputing,2009,72(13-15):3077-3084.
  • 5Cortes C,Mohri M,Rostamizadeh A.Two-stage learningkernel algorithms[C].Haifa,Israel:Proc of the 27th Inter-national Conference on Machine Learning,2010:239-246.
  • 6Wang G,Yeung D Y,Lochovsky F H.A kernel path algo-rithm for support vector machines[C].Corvalis,USA:Proc of the 24th International Conference on MachineLearning,2007:951-958.
  • 7WU K P,WANG S D.Choosing the kernel parameters forsupport vector machines by inter-cluster distance in the featurespace[J].Pattern Recognition,2009,42(5):710-717.
  • 8Lanckriet G R G,Bie T D,Cristianini N,et al.A statisticalframework for genomic data fusion[J].Bioinformatics,2004,20(16):2626-2635.
  • 9Bach F R,Lanckriet G R G,Jordan M I.Multiple kernellearning,conic duality,and the SMO algorithm[C].Banff,Canada:Proc of the 21st International Conference on MachineLearning,2004:41-48.
  • 10Sonnenburg S,Rtsch G,Schfer C,et al.Large scale mul-tiple kernel learning[J].Journal of Machine Learning Re-search,2006,7(1):1531-1565.

引证文献1

二级引证文献53

电子学报
2007年 第5期

相关作者

内容加载中请稍等...

相关机构

内容加载中请稍等...

相关主题

内容加载中请稍等...

浏览历史

内容加载中请稍等...
;
使用帮助 返回顶部