吴怀宇_中国科学院分享 http://blog.sciencenet.cn/u/wuhuaiyu 博士、副教授 「模式识别国家重点实验室」&「中国-欧洲信息,自动化与应用数学联合实验室」

博文

[CV论文读讲] Feature selection and extraction——mainly PCA

已有 2735 次阅读 2012-11-30 14:20 |系统分类:科研笔记| 论文, features, identify

相关下载详见 “视觉计算研究论坛”「SIGVC BBS」:http://www.sigvc.org/bbs/thread-36-1-2.html

Premise:  you have got a set of    
                features(measurement) of  samples.
•identify those variables that do not contribute to the classification task.


•find a transformation from the p measurements to a lower-dimensional feature space.

 


•select those d variables that contribute most to discrimination.


•Feature selection criteria: error rate, probabilistic distance, recursive calculation of separability measurement, criteria based on scatter matrices.

 


• PCA方法现阶段的应用?现阶段的计算能力上再讨论PCA还有没有意义?


•降维还有没有必要?

 


 



https://blog.sciencenet.cn/blog-4099-637813.html

上一篇:[CV论文读讲] The conjugate gradient method
下一篇:[CV论文读讲] Thin Plate Spline
收藏 IP: 123.118.210.*| 热度|

0

该博文允许注册用户评论 请点击登录 评论 (0 个评论)

数据加载中...
扫一扫,分享此博文

Archiver|手机版|科学网 ( 京ICP备07017567号-12 )

GMT+8, 2024-3-29 23:27

Powered by ScienceNet.cn

Copyright © 2007- 中国科学报社

返回顶部