非线性降维在高维医学数据处理中的应用

被引:14
作者
翁时锋
张长水
张学工
机构
[1] 清华大学自动化系
关键词
模式识别理论; 非线性降维; 乳腺癌; 肺癌;
D O I
10.16511/j.cnki.qhdxxb.2004.04.013
中图分类号
R311 [医用数学];
学科分类号
摘要
针对非线性高维医学数据降维的困难,引入了一种新的非线性降维方法Isomap,并从算法原理的角度讨论了方法在医学数据处理中的适用性。该文将Isomap应用在两个典型医学数据集(肺癌基因表达数据和乳腺癌病理数据)的分析中,发现它们的本质维数都低于3,因而可以得到在低维投影空间中的可视化表示。实验进一步将Isomap和主成份分析(PCA)的投影结果相比较,并统计类内距离,结果显示Isomap优于传统的线性降维技术。这说明了非线性降维技术在高维医学数据分析中的潜力。
引用
收藏
页码:485 / 488
页数:4
相关论文
共 7 条
  • [1] A global geometric framework for nonlinear dimensionality reduction. Tenenbaum J B,Silvam V D,Langford J C. Science . 2000
  • [2] Principal curves. Hastie T J,Stuetzle W. Journal of the American Statistical Association . 1989
  • [3] Multisurface method of pattern separation for medical diagnosis applied to breast cytology. Wolberg W H,Mangasarian O L. Proceedings of the National Academy of Sciences of the United States of America . 1990
  • [4] Classification of human lung carcinomas by mRNA expression profiling reveals distinct adenocarcinoma subclasses. Bhattacharjee A,Richards W G,Staunton J,et al. Proceedings of the National Academy of Sciences of the United States of America . 2001
  • [5] Robust linear programming discrimination of two linearly inseparable sets. Bennett K P,Mangasarian O L. O ptim Methed and Softw . 1992
  • [6] Cancer statistics. Greenlee R T,Hill -Harmon M B,Murray T,et al. CA A Cancer Journal for Clinicians . 2001
  • [7] Molecular classification of cancer: Class discovery and class prediction by gene expression monitoring. Golub T R,Slonim D K,Tamayo P,et al. Science . 1999