采用高维数据归约由药物判定方剂功效

Determing Function of Prescription According to Herb Virtue by A Reduce Method of High Dimensions

目的:探讨药物高维数据归约方法判断方剂功效的可行性。方法:以《中医类方辞典》的方剂为样本数据,将组方药物的数据量化,采用单纯的人工神经网络方法(ANN)和人工神经网络方法结合属性距离矩阵的高维数据方法(SARM)分别由药物判断方剂功效。结果与结论:采用人工神经网络方法并结合中医先验知识由药物判断客观方剂功效是可行的。

Objective： The article studies the virtue of prescription through the reduce method of herb virtue high dimensions. Method： The sample datum come from the book Similar Prescription of TCM. First it quantifies the property of the herb. Second it makes a comparison between Artificial Neural Network（ANN） and Similarity ANN based Reduction Method （SARM）. Results and Conclusion： It is feasible to conclude the virtue of prescription from the herb virtue by SARM.