基于支持向量回归机的中药类方配伍优化研究  

Study on the Compatibility Optimization of Categorized Formulas of Chinese Medicine Based on Support Vector Regression

作　　者：王秀凤[1] 张磊[1] 罗来成[1] 王建红[1] 

机构地区：[1]广东药学院基础学院,广州510006

出　　处：《生物医学工程学杂志》2012年第6期1058-1061,共4页Journal of Biomedical Engineering

基　　金：国家自然科学基金资助项目(30973977)

摘　　要：为了研究中药类方的配伍规律,将支持向量回归机(SVR)理论应用于类方药效预测的研究,建立了基于SVR的类方药效预测模型。根据此模型可以预测不同类方配伍的药效,对优化类方配伍及临床用药有一定的指导作用。将此模型应用于肾气丸、右归丸和右归饮3首补肾阳类方配伍规律的研究,通过模型预测得到多个配伍组的药效优于3首补肾阳类方的原方。Prediction on pharmacodynamic action of categorized formulas is presented with the theory of support vec tor regression （SVR） in this paper. A prediction model of pharmacodynamic action of categorized formulas based on SVR was set up in order to predict the law of the compatibility of the categorized formulas. Pharmaeodynamic action of various categorized formulas could be predicted based on this model. It is very significient to optimize the compati bility of categorized formulas and clinical practice. This model was applied to the research of the law of compatibility in three categorized formulas for tonifying kidney yang which contains shenqi pill, yougui pill and yougui drink. As indicated in the model prediction, pharmacodynamie actions of several compatibilities of the categorized formulas are superior to that of the three original formulas for tonifying kidney yang.

关 键 词：类方 配伍规律 支持向量回归机 预测 

分 类 号：R289.1[医药卫生—方剂学]