Optimization of extraction process of compound Clematidis Radix spray by support vector machine

L9 (3(4)) orthogonal experiment was used to design the extraction technology of compound Clematidis Radix spray. Weight coefficients of active ingredients and dry extract rate were solved by information entropy. Support vector machine (SVM) was established and the model parameters were optimized through the genetic algorithm. Grid search algorithm was used for optimization of extraction technology of Clematidis Radix spray. The optimal extraction technology was to extract Clematidis Radix spray in water with 6 times the weight of herbal medicine for 3 times, with 2 h once. Bias of value between real and predicted by SVM was 1.23%. SVM was compared with traditional intuitive analysis of orthogonal design. It indicates that the new method used to optimize the extraction parameters of compound Clematidis Radix spray is more accurate and reliable.