摘要
目的 对中药药性依其重要性进行更深层次的多值量化分析 ,将比定性的文字表述和简单的二值量化表示更有意义。方法 以补虚药为研究对象 ,药性为基本特征 ,在对其进行二值量化和多值量化的基础上 ,采用应用人工神经网络和决策树等数据库知识发现技术 ,进行对补虚药功效归类判别结果的影响研究。结论 结果表明药性的不同量化方法对补虚药的功效归类预测有一定影响 ,多值量化比二值量化具有更为理想的判别结果。
Objective To carry out a deeper level multivalued quantification analysis of TCM drug properties according their greatness of significance. Method Deficiency-nourishing drugs were taken as the study objects,and their TCM drug properties were taken as their basic characteristics. The TCM drug properties of deficiency-nourishing drugs were first treated by two-valued quantification and multivalued quantification,and then,techniques of knowledge discovery in database,such as the artificial neural network and decision tree,were used to find the influence of different quantification methods on the results of the effect attribution discrimination of deficiency-nourishing drugs. Results The multivalued quantification has a more ideal result of discrimination than the two-valued quantification. Conclusion The different methods for the quantification of TCM drug properties have certain influence on the prediction of the effect attribution of deficiency-nourishing drugs.
出处
《北京中医药大学学报》
CAS
CSCD
北大核心
2004年第4期7-9,共3页
Journal of Beijing University of Traditional Chinese Medicine
基金
国家自然科学基金资助课题 (No 3 0 3 71784)
关键词
中药药性
补虚药
人工神经网络
决策树
功效分类
量化
Deficiency-Nourishing Drugs
Artificial Neural Network
Decision Tree
Effect Attribution
Quantification