摘要
对不同地区的葛根类药材按药理活性强度进行正确分类.采用基于聚类的模式识别分类算法.采用基于聚类的模式识别分类算法,该分类算法将聚类技术引入有监督的自动分类并应用于中药材的模式识别,从而对葛根类中药的药理抗内毒素活性识别正确率达92%,优于经典k最近邻法与Bayers判别法.此方法有助于判定葛属植物的质量,可用于中药模式识别研究.
To correctly classify Pueraria medicines in different regions according to the intensity of pharmacological activity,the classification algorithm of pattern recognition based on clustering was used to introduce clustering technology into a supervised automatic classification and was applied to pattern recognition of Chinese herbal medicines.The classification algorithm＇s recognition activity was found to be 92% at Chinese Herbal Pueraria anti-endotoxin and better than the classical k nearest neighbor method and the Bayers Criterion.This method helps to determine the quality of Pueraria DC plants and may be used for pattern recognition of traditional Chinese medicine.
出处
《河北北方学院学报（自然科学版）》
2011年第1期48-51,共4页
Journal of Hebei North University:Natural Science Edition
基金
张家口市科学技术研究与发展项目(0921045B)
河北北方学院青年基金项目(Q2010008)
关键词
葛根
葛属
距离
准确率
模拟识别技术
root of Pueraria DC
Pueraria DC
distance
accuracy
pattern recognition technology
