RBF神经网络在中医药领域中的应用及其在SPSS 17.0软件中的实现  

Application of RBF-ANN in TCM and Establishment of RBF-ANN by SPSS 17. 0

作　　者：邹慧琴[1] 李硕[2] 陶欧[1] 颜素容[1] 林兆洲[1] 赵丽莹[1] 闫永红[1] 

机构地区：[1]北京中医药大学,北京100029 [2]北京交通大学,北京100044

出　　处：《中华中医药学刊》2015年第2期336-338,I0012,共4页Chinese Archives of Traditional Chinese Medicine

基　　金：国家自然科学基金项目(81403054)

摘　　要：径向基函数(Radial Basis Function,RBF)神经网络在中医药研究领域中得到广泛应用,包括中医舌脉象客观化研究[1]、方证分析、证候诊断、中药质量控制、指纹图谱研究、药材的真伪优劣鉴别[2]、中药活性成分药效的模拟与预测及构效关系研究、中药制剂全过程优化等。对近年来RBF神经网络在中医药领域的应用作简要介绍,同时运用实例就SPSS 17.0软件在RBF神经网络的设计和结果分析方面的应用作论述,实验结果进一步证实了利用RBF神经网络可建立优质、准确的判别模型,其操作简便、结果直观,可提高中医药研究者科研工作中对大批、多维数据的处理与分析效率;应用SPSS 17.0软件实现RBF神经网络的设计与建立具有推广和普及价值。Radial basis function article neural network（ RBF- ANN） has a wide range of application in traditional Chinese medicine（ TCM）,such as standardization and digitalization of tongue and pulse condition in clinical patients,relationship analyses between formulae and diseases or syndrome,research on fingerprints and identification of Chinese materia medica（ CMM）,prediction of active compounds in CMM and their relationship with CMM＇s quality. In this paper,the application of RBF- ANN in TCM was summarized and SPSS 17. 0 software was introduced. And samples from eight different species of Asteraceae family were purchased and analyzed by electronic nose. Based on their digital aroma data,RBF- ANN discriminative model was established in SPSS 17. 0 software and these different kinds of CMM samples were identified successfully in this RBF- ANN model. Experimental results indicate that RBF- ANN could be employed to construct a simple,accurate and reliable discriminative model,which surely helps researchers save their time in processing high dimensions data. This method has its promising value to be popularized.

关 键 词：中医药 人工神经网络 径向基函数 SPSS 17.0 

分 类 号：R222.19[医药卫生—中医基础理论]