摘要
目的:探讨模糊k-最近邻算法运用于葛根类药材模式识别的可行性。方法:选择6种化学成分的含量,对不同产地的多种葛根类中药的药理抗内毒素活性建立了模糊k-最近邻规则识别模式。结果:模糊k-最近邻规则对葛根类中药的药理抗内毒素活性识别正确率达100%,优于经典k-最近邻法与Bayers判别法。结论:模糊k-最近邻算法可用于中药模式识别研究。
Object： To explore the feasibility of Pattern Recognition of Pueraria DC. based on Fuzzy k- nearest neighbor algorithm. Methods. Pattern Recognition of Pueraria DC, from different localities in china was established by Fuzzy k-nearest neighbor algorithm for the classification of anti-endo-toxin activity with contents of six components inputs. Results： The validities of Pattern Recognition based on Fuzzy k-nearest neighbor algorithm was found to be 100%, which was better then that of KNN＇s and Bayers＇, Conclusion： Pattern Recognition of traditional Chinese medicine based on Fuzzy k-nearest neighbor algorithm was promis- ing.
出处
《数理医药学杂志》
2007年第6期839-841,共3页
Journal of Mathematical Medicine
关键词
模糊k-最近邻算法
K-最近邻算法
模式识别
葛属
Fuzzy k-nearest neighbor algorithm
k-nearest neighbor algorithm
pattern recognition
pueraria DC.
