Forecast of Ultrafiltration Flux of Traditional Chinese Medicine Volatile Oil-in-water Emulsion by Support Vector Machine Method

The ultrafiltration flux of traditional Chinese medicine volatile oil-in-water emulsion was forecasted by support vector machine(SVM) method.Forty groups' data were chosen to optimize parameters and train the model, by which to predict steady flux of the other ten groups' data.SVM method was compared with back-propagation neural net method.The results showed that the former method was superior to the latter in forecasting ultrafiltration flux.