一种基于核距离的核函数度量方法

计算机科学 ›› 2014, Vol. 41 ›› Issue (2): 72-75.

一种基于核距离的核函数度量方法

王裴岩,蔡东风

沈阳航空航天大学知识工程研究中心沈阳110136;沈阳航空航天大学知识工程研究中心沈阳110136

出版日期:2018-11-14 发布日期:2018-11-14
基金资助:
本文受国家自然科学基金(61073123),沈航校青年教师自选课题(201106Y)资助

Distance-based Kernel Evaluation Measure

WANG Pei-yan and CAI Dong-feng

Online:2018-11-14 Published:2018-11-14

摘要/Abstract

摘要： 核方法的效果依赖于所使用的核,因此核的选择和其参数的确定是至关重要的。从特定的数据中学习核需要核度量方法评价核的质量。核排列度量核与学习任务的一致性,因为它具有高效性和有效性,是目前应用最为广泛的核度量方法。然而,有研究表明,核排列仅是最优核函数的充分非必要条件。其主要原因是核排列在特征空间中不具有线性变换不变性。提出了一种新的核度量方法用于核选择,称其为核距离排列。该方法能够克服核排列的局限性,并且同样具有高效性和简单的形式。对比实验表明,该方法能够有效地对核进行度量。

关键词: 核方法,核度量,核距离中图法分类号TP391文献标识码A

Abstract: The success of kernel methods depends on the kernel,thus a choice of a kernel and proper setting of its parameters are of crucial importance．Learning a kernel from the data requires evaluation measures to assess the quality of the kernel．Recently,kernel target alignment (KTA),which measures the degree of agreement between a kernel and a learning task,has been widely used for kernel selection because of its effectiveness and efficiency．However,it is reported that KTA is only a sufficient condition to select a good kernel,but not a necessary condition．The reason is that KTA is not invariant under data translation in the feature space.This paper proposed a new measure for kernel selection named kernel distance target alignment (KDTA)．The measure not only overcomes the limitations of KTA but also possesses other properties like simplicity and efficiency．Comparative experiments indicate that the new measure is a good indication of the superiority of a kernel.

Key words: Kernel method,Kernel evaluation measure,Kernel distance

王裴岩,蔡东风. 一种基于核距离的核函数度量方法[J]. 计算机科学, 2014, 41(2): 72-75. https://doi.org/

WANG Pei-yan and CAI Dong-feng. Distance-based Kernel Evaluation Measure[J]. Computer Science, 2014, 41(2): 72-75. https://doi.org/

参考文献

[1] Schlkopf B,Smola A．Learning with Kernels [M]．MIT Press,Cambridge,Massachusetts,2002
[2] Gnen M,Alpaydin E．Multiple kernel learning algorithms [J]．Journal of Machine Learning Research,2011,12:2211-2268
[3] Girolami M,Rogers S．Hierarchic bayesian models for kernellearning [C]∥Proceedings of the 22nd internatinoal conference on machine learning．Bonn,Germany,Springer Verlag,August 2005:241-248
[4] Cheng S-O,Smola A J,Williamson R C．Learning the kernelwith hyperkernels [J]．Journal of Machine Learning Research,2005,6:1043-1071
[5] Cristianini N,Shawe-Taylor J,Elisseeff A,et al.On kernel-target alignment [J]．Advances in Neural Information Processing Systems,2001,4:367-373
[6] Baram Y．Learning by kernel polarization [J]．Neural Computation,2005,17:1264-1275
[7] Wang Ting-hua,Tian Sheng-feng,Huang Hou-kuan,et al．Lear-ning by local kernel polariz [J]．Neurocomputing,2009,72:3077-3084
[8] Nguyen C H,Ho Tu-bao．An efficient kernel matrix evaluation measure [J]．Pattern Recognition,2008,1:3366-3372
[9] Chudzian P．Evaluation measures for kernel optimization [J]．Pattern Recognition Letters,2012,33:1108-1116
[10] Schlkopf B．The kernel trick for distance [J]．Advances in Neural Information Processing Systems,2001,13:301-307
[11] Wang Ting-hua,Zhao Dong-yan, Tian Sheng-feng．An overview of kernel alignment and its applications [J]．Artificial Intelligence Review,November 2012
[12] Camargo J E,Gonz′alez F A．A multi-class kernel aligmentmethod for image collection summa- rization [C]∥Proceedings of the 14th Iberoamerican Conference on Pattern Recognition:Progress in Pattern Recognition,Image Analysis,Computer Vision,and Applications．Guadalajara,Mexico,Springer Verlag,November 2009:545-552
[13] Igel C,Glasmachers T,Mersch B,et al．Gradientbased optimization of kernel- target aligment for sequence kernels applied to bacterial gene start detections [J]．IEEE Transactions on Computational Biology and Bioinformatics,2007,4(2):216-226
[14] Wong W W L,Burkowski F J．Using kernel alignment to select features of molecular descriptors in a qsar study [J]．IEEE Transactions on Computational Biology and Bioinformatics,2011,8(5):1373-1384
[15] Ramona M,Richard G,David B．Multiclass feature selectionwith kernel gram matrix based criteria [J]．IEEE transactions on neural networks and learning systems,2012,23(10):1611-1623
[16] Hofmann T,Scholkopf B,Smola A J．Kernel methods in machinelearning [J]．The Annals of Statistics,2008,36(3):1171-1220
[17] Lesot M-J,Rifqi M．Similarity measures for binary and numerucal data:a survey [J]．International Journal of Knowledge Engineering and Soft Data Paradigms,2009,1(1):63-84

Metrics

Viewed

Full text

Abstract

Cited

Shared

Discussed