跳到主要导航 跳到搜索 跳到主要内容

Block-quantized kernel matrix for fast spectral embedding

  • Kai Zhang*
  • , James T. Kwok
  • *此作品的通讯作者
  • Hong Kong University of Science and Technology

科研成果: 书/报告/会议事项章节会议稿件同行评审

摘要

Eigendecomposition of kernel matrix is an indispensable procedure in many learning and vision tasks. However, the cubic complexity O(N3) is impractical for large problem, where N is the data size. In this paper, we propose an efficient approach to solve the eigendecomposition of the kernel matrix W. The idea is to approximate W with W̄ that is composed of m 2 constant blocks. The eigenvectors of W̄, which can be solved in O(m3) time, is then used to recover the eigenvectors of the original kernel matrix. The complexity of our method is only O(mN + m 3), which scales more favorably than state-of-the-art low rank approximation and sampling based approaches (O(m2N + m3)), and the approximation quality can be controlled conveniently. Our method demonstrates encouraging scaling behaviors in experiments of image segmentation (by spectral clustering) and kernel principal component analysis.

源语言英语
主期刊名ICML 2006 - Proceedings of the 23rd International Conference on Machine Learning
1097-1104
页数8
出版状态已出版 - 2006
已对外发布
活动ICML 2006: 23rd International Conference on Machine Learning - Pittsburgh, PA, 美国
期限: 25 6月 200629 6月 2006

出版系列

姓名ICML 2006 - Proceedings of the 23rd International Conference on Machine Learning
2006

会议

会议ICML 2006: 23rd International Conference on Machine Learning
国家/地区美国
Pittsburgh, PA
时期25/06/0629/06/06

指纹

探究 'Block-quantized kernel matrix for fast spectral embedding' 的科研主题。它们共同构成独一无二的指纹。

引用此