六味地黄丸缺味药的Bayes法和PRIMA法定性识别研究

Qualitative Recognition on Omitted Ingredients in Liuwei Dihuang Pills by Bayesian Analysis and PRIMA

目的：对六味地黄丸缺味药进行Ｂａｙｅｓ 法和ＰＲＩＭＡ法定性识别研究。方法：采用反相ＨＰＬＣ法对六味地黄丸缺味药模拟方的浸出物进行分析，选取９ 个色谱峰的峰面积与内标峰面积之比值作为样本特征变量，通过１６９ 个训练集样本建立了其中３ 种缺味药的Ｂａｙｅｓ 法和ＰＲＩＭＡ法判别分析数学模型。结果：３ 种缺味药４ 种模式的平均正确识别率Ｂａｙｅｓ 法和ＰＲＩＭＡ 法均为１００％ ，对１６９ 个预示集样本的平均预示率Ｂａｙｅｓ 法为１００％ ，ＰＲＩＭＡ法为９９ ．６％ 。结论：Ｂａｙｅｓ 法和ＰＲＩＭＡ 法能对六味地黄丸３ 种缺味药进行准确识别。

Objective: To qualitatively recognize the omitted ingredients in Liuwei Dihuang Pills by Bayesian analysis and PRIMA. Methods: The extracts of the pills were analyzed by HPLC, for each object, the areas of nine peaks relative to that of the internal standard were selected as classification features.The classification model was established on 169 training set objects.Result: The classification average accuracy was 100% for four patterns by Bayesian analysis,and 99.6% by PRIMA analysis. Conclusion: Bayesian analysis and PRIMA analysis can be used to classify and recognize the omitted ingredients in Liuwei Dihuang Pills.