A Novel Quality Classification Method to Measuring Chemical Contents in Tobacco Leaves

摘要：

Tobacco quality classification plays a significant role in its market price determination.Conventional methods including linear discriminant analysis, K-means clustering and BP-neural network flaw in capture the nonlinear structure. The use of support vector machine (SVM) has been shown to be a cost-effective technique. But it is used as a non-preprocessing way for a classification task. This paper extended SVM with kernel principal component analysis (KPCA) for extract valuable discriminatory information. The method is then applied to classify tobacco leaves quality of the Wulong country, one of the most important tobacco planting areas of Chongqing. The classification performance of the proposed method is proven superior compared with other statistical and machine learning methods.

关键词： Kernel principal component analysis Support vector machine Preprocessing Agriculture

作者: JIANG Wei LV Jiake

作者单位: College of Computer and Information Science, Southwest University, P.R.China, 400716

会议类型: 国际会议

会议名称: 2008 International Conference on IAEA(2008农业信息化、自动化与电气化国际会议)

会议地点: 镇江

会议语种:英文

页码: 191-196

在线出版日期: 2008-11-01（万方平台首次上网日期，不代表论文的发表时间）