|Table of Contents|

Using Optimization Method to Solve Linear Discriminant Equations

《南京理工大学学报》(自然科学版)[ISSN:1005-9830/CN:32-1397/N]

Issue:
2005年06期
Page:
745-747
Research Field:
Publishing date:

Info

Title:
Using Optimization Method to Solve Linear Discriminant Equations
Author(s):
SHAO LiYANG Yu-wang
Department of Computer Science and Technology,NUST,Nanjing 210094,China
Keywords:
optimization method pattern recognition linear discriminant analysis
PACS:
TP391.41
DOI:
-
Abstract:
The sequential quadrat ic programming ( SQP) method is adopted to solve the linear ( Fisher) discrim-i nant criterion model in the pattern recognition field. Several kinds of optimization techniques and algorithms are explained, and the high efficiency of SQP method compared to the ordinary solution methods for the linear discriminant criterion models is analyzed. With the help of the optimization toolbox in MATLAB, the SQP method is adopted to solve the Fisher discriminant equation for the two-class problems ( human face image) , which obtains satisfactory results. The research and implementation is a new attempt and contribution to the solut ion of Fisher discriminant model.

References:

[ 1] 孙即祥. 现代模式识别[M] . 长沙: 国防科技大学出版社, 2001.
[ 2] Liu Ke, Jiang Ren, Cheng Yongqing, et al. Color Human Face Image Recognition Based On StatisticalModel[ A] . Neural and Stochastic Methods in Image and Signal Processing [ A ] . SPIE, 1992, 1776: 702- 707.
[ 3] 金忠, 杨静宇, 陆建峰. 一种具有统计不相关的最佳鉴别矢量集[ J] . 计算机学报, 1999, 22( 10) : 1 105- 1 108.
[ 4] 解可新, 韩立兴. 最优化方法[M] . 天津: 天津大学出版社, 1997.
[ 5] 葛哲学. MATLAB 6. 5 辅助优化计算与设计[M] . 北京: 电子工业出版社, 2003.

Memo

Memo:
-
Last Update: 2013-03-03