七种无机元素与中药药性的相关性研究  

Study on the relationship of the properties of neutral medicine of traditional Chinese medicines and seven inorganic elements

作　　者：刘进[1] 文志宁[1] 覃洁萍[2] 王丽丽[2] 李梦龙[1] 

机构地区：[1]四川大学化学学院,四川成都610064 [2]广西中医学院,广西南宁530001

出　　处：《化学研究与应用》2009年第1期81-84,共4页Chemical Research and Application

基　　金：国家973计划资助项目(2007CD512602)

摘　　要：本文研究支持向量机用于中药药性识别的可行性。选择7种无机元素在中药中的含量为指标,运用支持向量机对193种不同药性的中药进行训练,建立平性与非平性中药的预测模型。结果训练集识别正确率95.03%,建立的模型对预测集中平性药的识别正确率为82.14%,对非平性药的识别正确率为70%,总正确率为73.58%。In this paper, the feasibility of application of support vector machines to Traditional Chinese Medicine （TCbf）property recognition was studied. The amount of 7 inorganic elements in TCM were selected as index, 193 kinds of TCM of different properties were trained through support vector machines so as to establish the neutral and non-neutral property of TCM prediction model. The recognition accuracy of training, set was 95.03%. Of the established model, the recognition accuracy for the neutral medicine in prediction set was 82.14% ,and 70% for non-neutral medicine, which made up an accuracy of 73.58% in total.

关 键 词：支持向量机 无机元素 中药 药性识别 

分 类 号：O652.9[理学—分析化学]