基于BP神经网络的中药性味归经与补虚药药效研究

基于BP神经网络构建76味补虚药分类模型。采用数据挖掘方法挖掘补虚药药效，利用Python语言实现BP神经网络，并构建药效分类模型及其应用。对12味补虚药进行测试，其中有10味预测结果正确，正确率为83.33%。通过药性分布可知，补虚药的主要特征为温、甘，归经为脾、肾、心经。采用中药性味归经为特征对中药分类进行预测，实验结果表明，基于BP神经网络的药效分类模型具有一定的可靠性和准确性，亦为中药分类方法研究提供了借鉴。

Study on Pharmacodynamics of TCM Meridian and Tonifying Deficiency Drugs Based on BP Neural Network

Based on the BP neural network， a classification model of 76 flavored drugs was established. Data mining method was used to mine the efficacy of tonic drugs， and BP neural network was used to construct the pharmacodynamic classification model and its application. 10 of the 12 flavours were tested correctly with a correct rate of 83.33%. Through the distribution of drug， we can see that the main characteristics of reinforcing deficiency medicine are warm and sweet， and the meridian is the spleen， kidney and heart meridian. In this study， the characteristics of flavor meridian of traditional Chinese medicine were used to predict the classification of traditional Chinese medicine. The experimental results show that the classification model based on BP neural network has certain reliability and accuracy， and provides a reference for the study of traditional Chinese medicine classification methods.