Grey relational analysis and LS-SVM modeling for the fingerprint-efficacy study of Yinhuang granules

摘要

In this article, the grey relational analysis method was used to identify the key constituents of Yinhuang granules according to the anti-respiratory syncytial virus activities in drug serum by in vitro laboratory experiments. Furthermore, a model that characterizes the relationship between constituents and median effective concentrations was established through the least squares support vector machine (LS-SVM) regression technique. The computational simulation showed that this model fitted well with the experimental data, and validation experimental results also supported the theoretical predictions.