面向田间籽棉成熟度判别的二种特征选择算法比较

王玲; 刘德营; 姬长英

doi:10.3788/OPE.20132108.2121

您当前的位置：

首页 >

文章列表页 >

面向田间籽棉成熟度判别的二种特征选择算法比较

信息科学 | 更新时间：2020-08-12

- 面向田间籽棉成熟度判别的二种特征选择算法比较
- Comparison of two Feature Selection Algorithms Oriented to Raw Cotton Ripeness Discrimination
- 光学精密工程 2013年21卷第8期页码：2121-2128
- 作者机构：
  
  南京农业大学工学院/江苏省现代设施农业技术与装备工程实验室,江苏南京,210031
- 作者简介：
- 基金信息：
  
  国家&ldquo();十五&rdquo();863计划资助项目();江苏省农机基金资助()
- DOI：10.3788/OPE.20132108.2121
  中图分类号： TP391.4
- 收稿日期：2013-02-06，
  
  修回日期：2013-04-06，
  
  网络出版日期：2013-08-20，
  
  纸质出版日期：2013-08-15
- 稿件说明：
移动端阅览
王玲刘德营姬长英. 面向田间籽棉成熟度判别的二种特征选择算法比较[J]. 光学精密工程, 2013,21(8): 2121-2128

WANG Ling LIU De-ying JI Chang-ying. Comparison of two Feature Selection Algorithms Oriented to Raw Cotton Ripeness Discrimination[J]. Editorial Office of Optics and Precision Engineering, 2013,21(8): 2121-2128
王玲刘德营姬长英. 面向田间籽棉成熟度判别的二种特征选择算法比较[J]. 光学精密工程, 2013,21(8): 2121-2128 DOI： 10.3788/OPE.20132108.2121.

WANG Ling LIU De-ying JI Chang-ying. Comparison of two Feature Selection Algorithms Oriented to Raw Cotton Ripeness Discrimination[J]. Editorial Office of Optics and Precision Engineering, 2013,21(8): 2121-2128 DOI： 10.3788/OPE.20132108.2121.

摘要

为了快速、准确地判别田间籽棉的成熟度，提取了描述棉瓣形状的15个结构特征，基于10折交叉验证比较了封装器下穷举搜索并基于封装器停止搜索（WE-W）和过滤器下启发式搜索并基于封装器停止搜索（FH-W）这二种特征选择算法的执行效率和分类性能。分别以验证集上Bayes分类器的误分率（WE-W）和训练集上的类可分性测量值（FH-W）为评价函数，在训练集上穷举搜索（WE-W）和启发式搜索（FH-W）最优l维特征子集，l=1

15，并于Bayes分类器在验证集上的平均误分率极小时停止搜索（WE-W和FH-W）。结果显示，WE-W和FH-W算法在预测集上于l=3处分别获得了85.39%（WE-W）和85.28%（FH-W）的平均识别率，表明FH-W算法执行效率高、分类性能好，对实际应用有参考意义。

Abstract

To discriminate the ripeness of cotton quickly and accurately

15 shape structure features were extracted from cotton images and the execute efficiencies and classification accuracy of their feature subset selection algorithms such as Wrapper-based Exhaustive searching and Wrapper-based stopping(WE-W) and Filter-based Heuristic searching and Wrapper-based stopping(FH-W) were compared by using 10-fold cross-validation. By taking the error rate of a Bayes classifier on validation set (WE-W) and the class-separability measuring value on a training set (FH-W)as assessing functions

the optimal l (l=1

15) feature subset was searched by using exhaustive (WE-W) and heuristic (FH-W) strategies on the training set

which stops at the minimum error rate of Bayes-classifier on the validation set(WE-W and FH-W). Experimental results show that the average classification rates of WE-W and FH-W algorithms on the prediction set are 85.39% (WE-W) and 85.28% (FH-W) at l=3

respectively. It concludes that the FH-W algorithm can be a reference in practice for its higher execute efficiency and good classification accuracy.

关键词

Keywords

references

孙伟艳. 模式分类中特征选择问题的研究[D]. 哈尔滨：哈尔滨理工大学，2009.SUN W Y. Research on feature selection for pattern classification [D]. Harbin:Harbin University of Science and Technology, 2009. (in Chinese)[2]NG W W Y, YEUNG D S, FIRTH M, et al.. Feature selection using localized generalization error for supervised classification problems using RBFNN [J]. Pattern Recognition, 2008, 41(12):3706-3719.[3]RUIZ R, RIQUELME J C, AGUILAR-RUIZ J S. Incremental wrapper-based gene selection from microarray data for cancer classification [J]. Pattern Recognition, 2006, 39(12):2383-2392.[4]LIANG J N, YANG S, WINSTANLEY A. Invariant optimal feature selection：A distance discriminant and feature ranking based solution [J]. Pattern Recognition, 2008(5),41:1429-1439.[5]LIU Y, ZHENG Y F. FS_SFS: A novel feature selection method for support vector machines [J]. Pattern Recognition, 2006, 39(7):1333-1345.[6]SEBBAN M, NOCK R. A hybrid filter/wrapper approach of feature selection using information theory [J]. Pattern Recognition, 2002, 35(4):835-846.[7]DONG M, KOTHARI R. Feature subset selection using a new definition of classifiability [J]. Pattern Recognition Letters, 2003, 24(9-10):1215-1225.[8]HUA J P, TEMBE W D, DOUGHERTY E R. Performance of feature-selection methods in the classification of high-dimension data [J]. Pattern Recognition, 2009, 42(3):409-424.[9]NAKARIYAKUL S, CASSASENT D P. An improvement on floating search algorithms for feature subset selection [J]. Pattern Recognition, 2009, 42(9):1932-1940.[10]李先锋，朱伟兴，纪滨，等. 基于图像处理和蚁群优化的形状特征选择与杂草识别[J]. 农业工程学报，2010, 26(10):178-182.LI X F, ZHU W X, JI B, et al.. Shape feature selection and weed recognition based on image processing and ant colony optimization [J]. Transactions of the CSAE, 2010, 26(10):178-182. (in Chinese)[11]董岩，张涛，李文明，等. 机载立体测绘相机滚转轴伺服系统的辨识与设计[J]. 光学精密工程，2011, 19(7): 1580-1587.DONG Y, ZHANG T, LI W M,et al.. Identification and design of roll axis servo system in airborne stereo mapping camera[J]. Opt. Precision Eng., 2011, 19(7): 1580-1587. (in Chinese)[12]董超，田联房. 最速上升关联向量机高光谱影像分类[J]. 光学精密工程，2012,20(6): 1398-1405.DONG CH, TIAN L F. Hyperspectral image classification by steepest ascent relevance vector machine[J]. Opt. Precision Eng., 2012,20(6):1398-1405. (in Chinese)[13]WEBB A R. 统计模式识别 [M]. 第二版. 王萍等译. 北京：电子工业出版社, 2004.WEBB A R. Statistical Pattern Recognition [M]. 2nd ed. Wang Ping et al. translation. Beijing：Publishing House of Electronics Industry, 2004. (in Chinese)[14]FRIEDMAN N, GEIGER D, GOLDSZMIDT M. Bayesian network classifiers [J]. Machine Learning, 1997, 29(2-3):131-163.[15]宋枫溪，高秀梅，刘树海，等. 统计模式识别中的维数削减与低损降维[J]. 计算机学报，2005, 28(11): 1915-1922.SONG F X, GAO X M, LIU SH H, et al.. Dimensionality reduction in statistical pattern recognition and low loss dimensional ity reduction [J]. Chinese Journal of Computers, 2005, 28(11): 1915-1922.(in Chinese)

浏览量

下载量

CSCD

文章被引用时，请邮件提醒。

提交

工具集

关联资源

多分支无锚框网络密集行人检测算法

基于监督保局子空间虚假近邻准则的原始特征选择