DocumentCode :
595071
Title :
Theoretical analysis of learning local anchors for classification
Author :
Junbiao Pang ; Qingming Huang ; Baocai Yin ; Lei Qin ; Dan Wang
Author_Institution :
Coll. of Comput. Sci. & Technol., Beijing Univ. of Technol., Beijing, China
fYear :
2012
fDate :
11-15 Nov. 2012
Firstpage :
1803
Lastpage :
1806
Abstract :
In this paper, we present a theoretical analysis on learning anchors for local coordinate coding (LCC), which is a method to model functions for data lying on non-linear manifolds. In our analysis several local coding schemes, i.e., orthogonal coordinate coding (OC-C), local Gaussian coding (LGC), local Student coding (LSC), are theoretically compared, in terms of the upper-bound locality error on any high-dimension data; this provides some insight to understand the local coding for classification tasks. We further give some interesting implications of our results, such as tradeoff between locality and approximation ability in learning anchors.
Keywords :
data models; encoding; pattern classification; LCC; LGC; LSC; OCC; approximation ability; classification tasks; data model functions; high-dimension data; learning local anchors; local Gaussian coding; local Student cod- ing; local coordinate coding; nonlinear manifolds; orthogonal coordinate coding; upper-bound locality error; Approximation methods; Educational institutions; Encoding; Error analysis; Manifolds; Upper bound; Vectors;
fLanguage :
English
Publisher :
ieee
Conference_Titel :
Pattern Recognition (ICPR), 2012 21st International Conference on
Conference_Location :
Tsukuba
ISSN :
1051-4651
Print_ISBN :
978-1-4673-2216-4
Type :
conf
Filename :
6460502
Link To Document :
بازگشت