基于混合Hausdorff距离的多示例学习近邻分类器

doi:10.6040/j.issn.1672-3961.1.2016.019

山东大学学报(工学版) ›› 2016, Vol. 46 ›› Issue (6): 15-22.doi: 10.6040/j.issn.1672-3961.1.2016.019

基于混合Hausdorff距离的多示例学习近邻分类器

陈泽华,尚晓慧,柴晶

太原理工大学信息工程学院, 山西太原 030024

收稿日期:2016-07-12 出版日期:2016-12-20 发布日期:2016-07-12
通讯作者: 柴晶(1983— ),男,山东聊城人,讲师,博士,主要研究方向为数据挖掘和机器学习.E-mail:jingchai@aliyun.com E-mail:zehuachen@163.com
作者简介:陈泽华(1974— ),女,山西太原人,副教授,博士,主要研究方向为智能信息处理.E-mail:zehuachen@163.com
基金资助:
国家自然科学基金资助项目(61403273,61402319);山西省自然科学基金资助项目(2014021022-4,2014021022-3)

Neighborhood related multiple-instance classifiers based on integrated Hausdorff distance

CHEN Zehua, SHANG Xiaohui, CHAI Jing

College of Information Engineering, Taiyuan University of Technology, Taiyuan 030024, Shanxi, China

Received:2016-07-12 Online:2016-12-20 Published:2016-07-12

摘要/Abstract

摘要： 通过对最小和最大Hausdorff距离的分析,提出混合Hausdorff距离将它们融合在一起以弥补任意单一Hausdorff距离的缺陷,并基于混合Hausdorff距离设计多示例学习近邻分类器。采用近邻分量分析模型能够优化混合Hausdorff距离中的权系数,从而得到在近邻分类准则下最优的混合Hausdorff距离。结果表明:相对于任意单一Hausdorff距离,基于混合Hausdorff距离的多示例学习近邻分类器通常能够获得更高的识别精度。

关键词: Hausdorff距离, 多示例学习, 近邻分量分析, 分类器, 权系数

Abstract: Based on the analysis of minimal Hausdorff(minH)and maximal Hausdorff(maxH)distances, integrated Hausdorff(intH)distance was proposed to combine minH and maxH, and used to design neighborhood related multiple-instance classifiers. The Neighborhood Component Analysis(NCA)model was used to learn the weighting coefficients in intH automatically and obtain the optimal intH according to the neighborhood related classification criterion. The experimental results showed that in most cases, compared with minH and maxH, intH could improve the classification accuracies of neighborhood related multiple-instance classifiers.

Key words: multiple-instance learning, weighting coefficients, Hausdorff distance, classifier, neighborhood component analysis

中图分类号:

TP181

陈泽华,尚晓慧,柴晶. 基于混合Hausdorff距离的多示例学习近邻分类器[J]. 山东大学学报(工学版), 2016, 46(6): 15-22.

CHEN Zehua, SHANG Xiaohui, CHAI Jing. Neighborhood related multiple-instance classifiers based on integrated Hausdorff distance[J]. JOURNAL OF SHANDONG UNIVERSITY (ENGINEERING SCIENCE), 2016, 46(6): 15-22.

参考文献

[1] 周志华. 多示例学习[M]. 知识科学中的基本问题研究. 北京:清华大学出版社, 2006:322-336.
[2] DIETTERICH T G, LATHROP R H, LOZANO-PEREZ T. Solving the multiple-instance problem with axis-parallel rectangles[J]. Artificial Intelligence, 1997, 89(1-2): 31-71.
[3] MARON O. Learning from ambiguity[D]. Boston: Massachusetts Institute of Technology, 1998.
[4] MARON O, LOZANO-PEREZ T. A framework for multiple-instance learning[C] //Advances in Neural Information Processing Systems 10. Denver, USA: MIT Press, 1998:570-576.
[5] ZHANG Q, GOLDMAN S A. EM-DD: an improved multiple-instance learning technique[C] //Advances in Neural Information Processing Systems 14. Vancouver, Canada: MIT Press, 2002:1073-1080.
[6] WANG J, ZUCKER J D. Solving the multiple-instance problem: a lazy learning approach[C] //Proceedings of the 17th International Conference on Machine Learning. Stanford, USA: Morgan Kaufmann Press, 2000:1119-1125.
[7] ANDREWS S, TSOCHANTARIDIS I, HOFMANN T. Support vector machines for multiple-instance learning[C] //Advances in Neural Information Processing Systems. Vancouver, Canada: MIT Press, 2003:561-568.
[8] GARTER T, FLACH P A, KOWALCZYK A, et al. Multi-instance kernels[C] //Proceedings of the 19th International Conference on Machine Learning. Sydney, Australia: Morgan Kaufmann Press, 2002:179-186.
[9] ZHOU Zhihua, SUN Yuyin, LI Yufeng. Multi-instance learning by treating instances as non-I.I.D. samples[C] //Proceedings of the 26th International Conference on Machine Learning. Montreal, Canada: ACM Press, 2009:1249-1256.
[10] LI W J, YEUNG D T. MILD: multiple-instance learning via disambiguation[J]. IEEE Transactions on Knowledge and Data Engineering, 2010, 22(1):76-89.
[11] SUN Y Y, NG M K, ZHOU Z H. Multi-instance dimensionality reduction[C] //Proceedings of the 24th AAAI Conference on Artificial Intelligence. Atlanta, USA: AAAI Press, 2010: 587-592.
[12] PING Wei, XU Ye, REN Kexin, et al. Non-i.i.d. multi-instance dimensionality reduction by learning a maximum bag margin subspace[C] //Proceedings of the 24th AAAI Conference on Artificial Intelligence Atlanta. USA: AAAI Press, 2010: 551-556.
[13] AMAR R A, DOOLY D R, GOLDMAN S A, et al. Multiple-instance learning of real-valued data[C] //Proceedings of the 18th International Conference on Machine Learning. Williamstown, USA: Morgan Kaufmann Press, 2001: 3-10.
[14] 薛晓冰,韩洁凌,姜远,等. 基于多示例学习技术的 Web 目录页面链接推荐[J].计算机研究与发展, 2007,44(3):406-411. XUE Xiaobing, HAN Jieling, JIANG Yuan, et al. Link recommendation in web index page based on multi-instance learning techniques[J]. Journal of Computer Research and Development, 2007, 44(3):406-411.
[15] ROWEIS S, HINTON G, SALAKHUTDINOV R. Neighbourhood component analysis[C] //Advances in Neural Information Processing Systems. Vancouver, Canada: MIT Press, 2005:513-520.
[16] PRESS W H, TEUKOLSKY S A, VETTERLING W T, et al. Numerical recipes: the art of scientific computing[M]. New York: Cambridge University Press, 2007.
[17] ZHANG Minling, ZHOU Zhihua. Multi-instance clustering with applications to multi-instance prediction[J]. Applied Intelligence, 2009, 31(1):47-68.
[18] ZAFRA A, PECHENIZKIY M, VENTURA S. ReliefF-MI: an extension of ReliefF to multiple instance learning[J]. Neurocomputing, 2012, 75(1):210-218.

多维度评价

Viewed

Full text

Abstract

Cited

Shared

Discussed

基于混合Hausdorff距离的多示例学习近邻分类器

Neighborhood related multiple-instance classifiers based on integrated Hausdorff distance

PDF (PC)

摘要/Abstract

引用本文

使用本文

参考文献

相关文章 11

多维度评价

本文评价

推荐阅读 10

[1]	侯霄雄,许新征,朱炯,郭燕燕. 基于AlexNet和集成分类器的乳腺癌计算机辅助诊断方法[J]. 山东大学学报 (工学版), 2019, 49(2): 74-79.
[2]	张红斌,邱蝶蝶,邬任重,朱涛,滑瑾,姬东鸿. 基于极端梯度提升树算法的图像属性标注[J]. 山东大学学报 (工学版), 2019, 49(2): 8-16.
[3]	于曰伟,周长城,赵雷雷,邢玉清,石沛林. 基于交替迭代的车辆主动悬架LQG控制器设计[J]. 山东大学学报(工学版), 2017, 47(4): 50-58.
[4]	曾碧, 毛勤. 改进的室内三维模糊位置指纹定位算法[J]. 山东大学学报(工学版), 2015, 45(3): 22-27.
[5]	江伟坚1,2,郭躬德1,2*,赖智铭1,2. 基于新Haar-like特征的Adaboost人脸检测算法[J]. 山东大学学报(工学版), 2014, 44(2): 43-48.
[6]	赵加敏,冯爱民*,刘学军. 局部密度嵌入的结构单类支持向量机[J]. 山东大学学报(工学版), 2012, 42(4): 13-18.
[7]	李玉鑑,孟东霞*,桂智明. 几何集成的改进——特征边界点快速计算[J]. 山东大学学报(工学版), 2011, 41(4): 56-60.
[8]	郭剑毅1,2,雷春雅1,余正涛1,2,苏磊1,2,赵君1,田维1. 基于信息熵的半监督领域实体关系抽取研究[J]. 山东大学学报(工学版), 2011, 41(4): 7-12.
[9]	冯爱民1,刘学军1,陈斌2. 结构大间隔单类分类器[J]. 山东大学学报(工学版), 2010, 40(3): 6-12.
[10]	张丽梅1,2 ,乔立山1,2 ,陈松灿1 . 基于张量模式的特征提取及分类器设计综述[J]. 山东大学学报(工学版), 2009, 39(1): 6-14.
[11]	陈涛,方志刚,徐洁 . 基于人脸和语音的混合型身份认证系统[J]. 山东大学学报(工学版), 2008, 38(2): 56-60 .